Determining the quality of probabilistic safety assessment (PSA) for

IAEA-TECDOC-1511

Determining the quality ofprobabilistic safety assessment

(PSA) for applications innuclear power plants

July 2006

IAEA SAFETY RELATED PUBLICATIONS

IAEA SAFETY STANDARDS

Under the terms of Article III of its Statute, the IAEA is authorized to establish or adopt standards of safety for protection of health and minimization of danger to life and property, and to provide for the application of these standards.

The publications by means of which the IAEA establishes standards are issued in the IAEA Safety Standards Series. This series covers nuclear safety, radiation safety, transport safety and waste safety, and also general safety (i.e. all these areas of safety). The publication categories in the series are Safety Fundamentals, Safety Requirementsand Safety Guides.

Safety standards are coded according to their coverage: nuclear safety (NS), radiation safety (RS), transport safety (TS), waste safety (WS) and general safety (GS).

Information on the IAEA’s safety standards programme is available at the IAEA Internet site

http://www-ns.iaea.org/standards/

The site provides the texts in English of published and draft safety standards. The texts of safety standards issued in Arabic, Chinese, French, Russian and Spanish, the IAEA Safety Glossary and a status report for safety standards under development are also available. For further information, please contact the IAEA at P.O. Box 100, A-1400 Vienna, Austria.

All users of IAEA safety standards are invited to inform the IAEA of experience in their use (e.g. as a basis for national regulations, for safety reviews and for training courses) for the purpose of ensuring that they continue to meet users’ needs. Information may be provided via the IAEA Internet site or by post, as above, or by e-mail to [email protected].

OTHER SAFETY RELATED PUBLICATIONS

The IAEA provides for the application of the standards and, under the terms of Articles III and VIII.C of its Statute, makes available and fosters the exchange of information relating to peaceful nuclear activities and serves as an intermediary among its Member States for this purpose.

Reports on safety and protection in nuclear activities are issued in other publications series, in particular the Safety Reports Series. Safety Reports provide practical examples and detailed methods that can be used in support of the safety standards. Other IAEA series of safety related publications are the Provision for the Application of Safety Standards Series, the Radiological Assessment Reports Series and the International Nuclear Safety Group’s INSAG Series. The IAEA also issues reports on radiological accidents and other special publications.

Safety related publications are also issued in the Technical Reports Series, the IAEA-TECDOC Series, the Training Course Series and the IAEA Services Series, and as Practical Radiation Safety Manuals and Practical Radiation Technical Manuals.Security related publications are issued in the IAEA Nuclear Security Series.

IAEA-TECDOC-1511

Determining the quality ofprobabilistic safety assessment

(PSA) for applications innuclear power plants

July 2006

The originating Section of this publication in the IAEA was:

Safety Assessment Section International Atomic Energy Agency

Wagramer Strasse 5 P.O. Box 100

A-1400 Vienna, Austria

DETERMINING THE QUALITY OF PROBABILISTIC SAFETY ASSESSMENT (PSA) FOR APPLICATIONS IN NUCLEAR POWER PLANTS

IAEA, VIENNA, 2006 IAEA-TECDOC-1511 ISBN 92–0–108706–3

ISSN 1011–4289 © IAEA, 2006

Printed by the IAEA in Austria July 2006

FOREWORD Probabilistic safety assessment (PSA) of nuclear power plants (NPPs) complements

the traditional deterministic analysis and is widely recognized as a comprehensive, structured approach to identifying accident scenarios and deriving numerical estimates of risks dealing with NPP operation and associated plant vulnerabilities. Increasingly, during the last years, PSA has been broadly applied to support numerous applications and risk-informed decisions on various operational and regulatory matters. The expanded use of PSA in the risk-informed decision making process requires that PSA possess certain features to ensure its technical consistency and quality.

This publication is aimed to further promote the use and application of PSA

techniques in Member States. The publication provides a comprehensive list of PSA applications and describes what technical features (termed ‘attributes’) of a PSA should be satisfied to reliably support the PSA applications of interest. A consideration has been also given to the basic set of attributes characterizing a ‘base case PSA’ that is performed with the purpose of assessing the overall plant safety.

This publication can support PSA practitioners in appropriate planning of a PSA

project taking into account possible uses of the PSA in the future. The publication can be also used by reviewers as an aid in assessing the quality of PSAs and judging the adequacy of a PSA for particular applications. In addition, it is also foreseen to use the publication to support the independent peer reviews conducted by the IAEA in the framework of the International Probabilistic Safety Assessment Review Team (IPSART) safety service.

The IAEA acknowledges the work of all of the participating experts and wishes to

thank them for their valuable contribution to this publication. The IAEA would like to especially thank K. Fleming of Karl N. Fleming Consulting Services (USA), R. Gubler of Ingburo Dr. Reinhard Gubler (Switzerland), P. Hellstrom of Relcon (Sweden), A. Lyubarskiy of SEC NRS (Russian Federation), G. Parry of the US NRC (USA), and G. Schoen and R. Schultz of HSK (Switzerland), for the active support and effective contribution to the IAEA project on development of this TECDOC.

The IAEA officer responsible for the preparation of this publication was I. Kouzmina

of the Division of Nuclear Installation Safety.

EDITORIAL NOTE

The use of particular designations of countries or territories does not imply any judgement by the publisher, the IAEA, as to the legal status of such countries or territories, of their authorities and institutions or of the delimitation of their boundaries.

The mention of names of specific companies or products (whether or not indicated as registered) does not imply any intention to infringe proprietary rights, nor should it be construed as an endorsement or recommendation on the part of the IAEA.

CONTENTS

1. INTRODUCTION.................................................................................................................. 1 1.1. Background ................................................................................................................... 1 1.2. Objectives of the report................................................................................................. 2 1.3. Quality of a PSA for an application.............................................................................. 3 1.4. Scope of the report ........................................................................................................ 5 1.5. Structure of the report ................................................................................................... 6 1.6. Applicability ................................................................................................................. 6

1.6.1. Applicability for reactors other than vessel type LWRs .................................... 6 1.6.2. Applicability for NPPs in different stages of the plant’s lifetime...................... 7

2. OVERVIEW OF PSA APPLICATIONS............................................................................... 9 2.1. PSA application categorization..................................................................................... 9 2.2. PSA results and metrics used in decision making ........................................................ 9

3. PROCEDURE TO ACHIEVE QUALITY IN PSA APPLICATIONS ............................... 12 3.1. PSA elements and attributes ....................................................................................... 12 3.2. Coding scheme for attributes identifiers ..................................................................... 13 3.3. Connection to IAEA PSA guidelines.......................................................................... 14 3.4. The procedure for use of the TECDOC ...................................................................... 14

4. PSA ELEMENT ‘IE’: INITIATING EVENTS ANALYSIS .............................................. 17 4.1. Main objectives ........................................................................................................... 17 4.2. Initiating events analysis tasks and their attributes..................................................... 18

5. PSA ELEMENT ‘AS’: ACCIDENT SEQUENCE ANALYSIS ......................................... 35 5.1. Main objectives ........................................................................................................... 35 5.2. Accident sequence analysis tasks and their attributes................................................. 35

6. PSA ELEMENT ‘SC’: SUCCESS CRITERIA FORMULATION AND SUPPORTING ANALYSIS................................................................................................ 44

6.1. Main objectives ........................................................................................................... 44 6.2. Success criteria formulation and supporting analysis tasks and their attributes ......... 44

7. PSA ELEMENT ‘SY’: SYSTEMS ANALYSIS ................................................................. 52 7.1 Main objectives ........................................................................................................... 52 7.2. Systems analysis tasks and their attributes ................................................................. 52

8. PSA ELEMENT ‘HR’: HUMAN RELIABILITY ANALYSIS.......................................... 66 8.1. Main objectives ........................................................................................................... 66 8.2. Human reliability analysis tasks and their attributes .................................................. 66

9. PSA ELEMENT ‘DA’: DATA ANALYSIS ....................................................................... 81 9.1. Main objectives ........................................................................................................... 81 9.2. Data analysis tasks and their attributes ....................................................................... 81

10. PSA ELEMENT ‘DF’: DEPENDENT FAILURES ANALYSIS...................................... 95 10.1. Main objectives ........................................................................................................... 95 10.2. Dependent failure analysis tasks and their attributes .................................................. 96

11. PSA ELEMENT ‘MQ’: MODEL INTEGRATION AND CDF QUANTIFICATION... 105 11.1. Main objectives ......................................................................................................... 105 11.2. Model integration and CDF quantification tasks and their attributes ....................... 105

12. PSA ELEMENT ‘RI’: RESULTS ANALYSIS AND INTERPRETATION .................. 112 12.1. Main objectives ......................................................................................................... 112 12.2. Results analysis and interpretation tasks and their attributes.................................... 112

13. DETERMINATION OF SPECIAL ATTRIBUTES FOR PSA APPLICATIONS ......... 117

14. CONCLUSIONS.............................................................................................................. 135

APPENDIX I: RISK METRIC DEFINITIONS..................................................................... 137

APPENDIX II: PSA APPLICATIONS.................................................................................. 143

APPENDIX III: MAPPING THE PSA ELEMENTS TO THE PSA TASKS ADDRESSED IN THE IAEA PSA GUIDELINES........................................................... 161

REFERENCES....................................................................................................................... 163

ABBREVIATIONS................................................................................................................ 165

CONTRIBUTORS TO DRAFTING AND REVIEW ........................................................... 169

1. INTRODUCTION 1.1. Background

To date, probabilistic safety assessments (PSAs) have been performed for the vast

majority of nuclear power plants (NPPs) worldwide and are under various stages of development for most of the remaining NPPs. PSA provides a comprehensive, structured approach to identifying accident scenarios and deriving numerical estimates of risks. In addition to the traditional deterministic analysis, it is a powerful tool for identification of significant accident sequences1 and associated plant vulnerabilities dealing with the design and operation of the plant. General guidance on performance and independent verification of the safety assessments for NPPs, both deterministic and probabilistic, is provided in the IAEA Safety Standards, e.g. in Ref. [1].

PSA is increasingly being used in many countries, in a complementary manner to the

traditional deterministic analysis and defence-in-depth considerations, as part of the decision making process to assess the level of safety of nuclear power plants and to support various risk-informed applications. Regulatory bodies in many countries require that a PSA be performed for licensing purposes. PSA has reached the point where, if performed to acceptable standards, it can considerably influence the design and operation of nuclear power plants. The quality of PSA is then becoming a matter of the ‘robustness’ of the decisions.

In order to promote the use and application of PSA techniques in Member States, the

IAEA has developed detailed technical guidance on how to carry out PSA for nuclear power plants. There are publications describing the overall process and procedures of performing a PSA (see Refs [2-4]). For specific PSA areas or tasks where it was felt that more detailed guidance is needed, separate publications have been produced, such as for common cause failure (CCF) modelling (see Ref. [5]) or human reliability analysis (HRA) (see Ref. [6]). The publications provide information and recommendations and reflect accepted practices consistent with the knowledge at the time they were written. Reference [2], for example, provides procedures for conducting Level-1 PSAs for internal events for full power initial conditions in accordance with the state of the art of PSA in the beginning of nineties. That publication has been successful in helping to standardize the framework, terminology, content and format of documentation of PSAs in IAEA Member States, while providing for flexibility to introduce new and alternative methods.

Increasingly, during the last years PSA has been broadly applied to support numerous

applications, such as risk-informed changes to technical specifications, risk-based plant configuration control, maintenance program optimization, etc. The IAEA has developed a technical document (see Ref. [7]), which summarizes information on up-to-date PSA applications and includes technical and methodological aspects, examples, and limitations, as well as the regulatory perspective on the use of PSA and numerical goals and acceptance criteria for decision making. A number of applications require specific features of the PSA and of certain PSA elements. A Technical Committee meeting, held in Vienna, May 28–June 1, 2001, on quality and consistency of PSAs identified the need for guidance on a

1 In this publication, it is assumed that the user will define an objective criterion to identify what is a significant accident sequence. An example of such a criterion is the following: a significant accident sequence is one of the set of sequences, defined at the functional or systemic level, that, when ranked in decreasing order of frequency, comprise a significant percentage (e.g. 95%) of the core damage frequency (CDF), or that individually contributes more than a measurable percentage (e.g. 1%) to CDF.

1

process to review a PSA to determine its technical adequacy for addressing specific applications, or in other words, to provide guidance to assure that a PSA is of sufficient quality to support the application. The same idea was highlighted at the Conference on Topical Issues in Nuclear, Radiation and Radioactive Waste Safety (IAEA, September 2001) that emphasized the necessity to ensure a high quality of PSAs for the support of risk-informed decision making. The meaning of ‘high quality’ in this context was meant to be different for each PSA and to be defined as being commensurate with the intended use of a PSA. The present publication is a response to those recommendations.

The publication takes into consideration the advanced worldwide experience in the

area of PSA quality assessment and verification, and in particular the ASME Standard for Probabilistic Risk Assessment for Nuclear Power Plant Applications (Ref. [8]). The starting point for the development of the technical attributes presented here in Sections 4 through 12 as the basis for the assessment of technical adequacy of a PSA was the set of requirements for the Capability Category II PSA presented in the ASME Standard (Capability Category II requirements are representative of currently accepted good industry practices in the USA.)

1.2. Objectives of the report

Various applications of PSA require that PSAs used to support those applications have

certain characteristics in terms of their scope, degree of detail, technical adequacy of the modelling, the capability and flexibility to perform the required calculations, the capability to support interpretation of the results, the quality and type of the data used, and of the assumptions made in modelling important aspects. The features of a PSA that are necessary to support specific applications vary with the application. This report provides information regarding the features, written in the form of attributes of the major PSA elements, which are appropriate for carrying out various PSA applications. In so doing, this publication provides a basis for judging the quality of the PSA used to support an application as discussed in the next section. General attributes are formulated for a ‘base case PSA’ that in the framework of this publication is defined to be that PSA that is used to assess the overall plant safety level. Special attributes are provided for specific applications where appropriate.

The notion of ‘PSA quality’ should be distinguished from the notion of ‘quality

assurance’. ‘PSA quality’ for a specific purpose refers to the technical adequacy of the methods, level of detail and data used to develop the PSA model. In order to assure that the chosen methods and data are used, applied, and documented in an adequate and controlled manner, a dedicated quality assurance programme needs to be established that also addresses applications of PSA. How to set up and effectively apply an appropriate quality assurance programme for PSA and its applications is described in the publication ‘A framework for a quality assurance programme for PSA’, IAEA-TECDOC-1101, Ref. [9]. As distinct from that publication, the present TECDOC focuses on the technical information regarding approaches, methodology and data to obtain appropriate technical PSA features for specific applications. Thus, for a specific PSA application with a particular PSA, the approach provided in the publication can be used as a basis to formulate a specific technical framework for carrying out the PSA application. For these reasons this publication concentrates on technical PSA aspects. In Figure 1 the overall framework for the assurance of the quality of PSA results for applications is shown identifying the roles of the existing IAEA publications and the present publication.

It is expected also that the publication will provide a technical framework for the PSA-related services and International Probabilistic Safety Assessment Review Team (IPSART)

2

missions being conducted by the IAEA on request of Member States, in addition to the existing guidelines, i.e. Ref. [10].

1.3. Quality of a PSA for an application

For the purposes of this publication, PSA quality is defined in general terms in the

following way: “In the context of an application, the PSA is of an appropriate quality if it conforms to a set of attributes that are appropriate for the application.”

Fig. 1. Framework of PSA quality and supporting IAEA publications. The key to defining quality then is in the definition of the attributes. The attributes that

are required for a particular application depend on the purpose and characteristics of the application. When used as an input to a decision, the attributes required are a function of the process for decision making, and in particular address the acceptance criteria or guidelines with which the PSA results are to be compared. The acceptance criteria are generally in the form of a numerical value associated with a specific metric. Examples of metrics are the absolute value of, or increase in, core damage frequency, and importance measures. The metrics commonly used are defined in Appendix I to this publication. The PSA has to be capable of evaluating the appropriate metrics for each application. However, the method for performing the comparison of the results of the PSA with the criterion also has an impact on the attributes required. For example, the criterion may require use only of a mean value, or it may require the full characterization of uncertainty as a probability distribution on the value of the metric. Thus identifying the additional attributes requires an understanding of the method proposed for generating and using the PSA results.

IAEA-TECDOC DETERMINING THE

QUALITY OF PSA FOR APPLICATIONS IN

NPPs

IAEA-TECDOC-1101 A Framework for a Quality Assurance Program for PSA

Technical Quality Attributes are defined

and used

The work on PSA is organized and performed

in such a way that required technical

attributes are met without introducing mistakes and

errors

The documentation is performed in a

manner that supports review and future

upgrades

Safety Series No. 50-C/SG-Q

Quality Assurance for Safety in Nuclear Power Plants and other Nuclear

Installations

Provision of the Quality of PSA Results for Applications

Safety Guide No. NS-G-1.2

Safety Assessment and Verification for Nuclear

Power Plants

3

Two types of attributes are defined in this publication:

- General attributes, which apply for a typical ‘base case PSA’ (for the definition of a ‘base case PSA’ see the discussion below). The general attributes apply for all PSAs and applications.

- Special attributes, which generally provide enhanced capabilities supporting certain applications of a PSA. Special attributes may not be met in a ’base case PSA’.

The purpose of a ‘base case PSA’ is the assessment of the overall plant safety as

described for example in Appendix II of this publication. Thus, the set of general attributes describing the technical features of a ‘base case PSA’ in this publication corresponds to the PSA application ‘Assessment of the Overall Plant Safety’. The general attributes represent a fundamental set of attributes that can be recognized as being associated with the performance of a technically correct PSA in accordance with the present state of the art methodology and technology. According to Ref. [10], “the current state of the art of PSA is defined by the way PSAs have been practically performed in recent years by Member States according to existing guidelines and using accepted methodologies and techniques.” To summarize, it is understood in this publication that the general attributes represent a minimum set of the attributes needed to perform a state of the art PSA with the aim to assess the overall plant safety. State-of-the-art is taken to be synonymous with generally accepted good practice.

Special attributes provide elevated capabilities in terms of resolution, specificity,

scope, realism, and less uncertainty for aspects of the PSA needed to support specific applications, but still corresponding to the current state of the art. Special attributes are defined in such a way that, when they are met, the corresponding general attributes will certainly be met. Different PSA applications may require different special attributes.

Special attributes may arise because of the need to model specific impacts of changes

proposed by the application, which may require a higher level of detail for certain elements than required for the base case as defined in this publication. In addition, special attributes may be required to address unique acceptance criteria for the application.

On the other hand, there might be applications for which not all the attributes would need to be met, or for which some attributes can be relaxed. These are applications for which either the risk information required is limited, or for which the approach to decision making compensates for a lesser level of detail or plant-specific fidelity in the PSA by making a more conservative decision than would be the case for the more detailed, plant-specific model. An example of the latter is an application that addresses relaxation of requirements on components considered to be of low safety significance. Use of a more detailed and more plant-specific PSA would allow more components to be classified as low safety significant, when compared with what would result from use of a less detailed model. However, even in this case, the PSA used to support that application must be technically adequate.

For many applications, the acceptance criteria may require the consideration of all

contributors to risk. It is recognized that specialized PSA methods are needed to perform the analysis of core damage resulting from internal hazards, such as internal fires and floods, or external hazards, such as earthquakes, high winds, etc., and from different plant operating modes, such as low power and shutdown modes. These specialized analyses are identified in

4

this publication as being separate modules2 of a PSA. The scope of a PSA is defined by the modules it contains.

Which attributes are met determines to some extent the role the PSA can play in the

decision making process. When it is clear that the confidence in the accuracy of the PSA results is high, the PSA can play a significant role. When confidence in the accuracy is less, it must play a lesser role. However, in either case, the PSA still has to have a quality commensurate with its role. What makes the distinction between these cases is that those attributes that enhance realism are not necessary met in the latter case.

When using information presented in this publication, it is proposed that in case a PSA

analyst considers that an application does not necessarily require compliance with a general or special attribute or attributes, this should be reliably justified in terms of the analysis consistency and absence of impact of a missing attribute(s) on PSA results and insights used for decision making. 1.4. Scope of the report

The detailed IAEA PSA procedures mentioned above (Refs [2-4]) mainly concentrate

on general features and content of PSAs. In these publications, a limited consideration is given to the particular features of PSA conditioned by specific PSA applications. It should be also mentioned that a number of approaches and techniques described in these procedures, in particular in the Level-1 PSA procedure (see Ref. [2]), have been further developed, so the present publication takes into account the current state of the art regarding various aspects related to PSA methodologies.

In recognition of the different levels of maturity in the state of the art for the various

PSA modules, and due to a comprehensive amount of information to be covered, the scope of this publication is restricted to a Level-1 PSA for at power operation for internal events caused by random equipment failures and operator errors. Consideration is not given to other sources of radioactivity except for the reactor core. Level-2 PSA, internal fires and floods, external hazards like earthquakes, tornadoes, and other natural and man-induced hazards are not included in the scope of this publication, and neither is PSA for the shutdown and low power operation modes. These PSA modules could be covered in separate publications later. However, it should be specifically pointed out that applications may require that the scope of the PSA is complete in terms of consideration of all relevant contributors to plant risk and analysis levels. It is not the intent of this publication to address what has to be done to compensate for the limited scope of the PSA in these circumstances. Nor does this publication attempt to describe what has to be done to compensate for attributes that are not met. These considerations are left to the decision-makers. However, Appendix II provides a general discussion regarding what PSA scope and risk metrics may be needed for specific applications.

An emphasis is made on describing the attributes of a ‘base case PSA’ being

fundamental for other considerations relating to specific PSA applications. The publication concentrates also on describing the appropriate features and attributes of PSA and of PSA elements and relates them to specific applications by indicating additional features and characteristics important from the viewpoint of specific applications. Only a summary of PSA 2 A PSA module is a probabilistic safety analysis, which addresses a certain type of hazard (e.g. high winds, earthquakes, internal fires), plant operating mode (full power, low power, shutdown), radioactivity source (reactor core, spent fuel pool), and analysis level (Level-1, Level-2, Level-3).

5

approaches, techniques, and tasks is given. The publication provides information on what has to be done rather than how it should be done. Thus, regarding detailed procedures for PSA tasks, reference is made to the appropriate available PSA procedures and the publication is not intended to replace them. 1.5. Structure of the report

Because this report is oriented towards supporting applications of PSA, first, an

overview of current applications is given in Section 2. Section 3 introduces the main PSA elements, and provides a description of the process one should follow to determine whether the PSA is of an appropriate quality for an application of interest. The attributes of the PSA elements are provided in Sections 4 through 12 separately for each PSA element, covering both general attributes (applicable for the ‘base case PSA’), and application-specific ones (i.e. special attributes). Section 13 discusses the special attributes appropriate for PSA applications and outlines a practical procedure for determination of the special attributes relevant for the application of interest. It also provides a table mapping the special attributes to the PSA applications. Conclusions are provided in Section 14. Appendix I provides definitions of the risk metrics referred to in the publication. Appendix II provides summary information on PSA applications, including their general description, applicable risk metrics, remarks on use of PSA models to support specific applications, and examples. Appendix III presents a table linking the PSA elements discussed in Sections 4 through 12 to the list of PSA tasks from the IAEA Procedure Guide on Level-1 Internal Event PSA (see Ref. [2]).

1.6. Applicability

There are three major limitations regarding the applicability of this publication which are as follows: 1. The information presented is directed towards PSA and PSA applications for nuclear

power plants. Thus, this publication is not directly applicable for research reactors, for example.

2. The publication focuses on PSA and PSA attributes for vessel type light water reactors

(LWRs), although a vast majority of general and special attributes are applicable for other reactor types as well. The applicability of the PSA element descriptions and of PSA attributes given in this publication for nuclear power plants with other reactor types is discussed below.

3. The publication is focused on PSA approaches, modelling and data for a typical ‘mature’

nuclear power plant, which has been in operation for a number of years without major changes in the plant. The applicability of the PSA elements descriptions and of PSA attributes given in this publication for nuclear power plants in other stages of the plants lifetime is discussed below.

1.6.1. Applicability for reactors other than vessel type LWRs

The present predominant reactor types for nuclear power plants are vessel type LWRs. PSA approaches and techniques have therefore been mostly developed and applied for this kind of NPPs. For this reason the publication focuses on PSA and PSA attributes for vessel type LWRs. Most of the PSA approaches and techniques can also be applied and used for other reactor types such as gas cooled reactors and CANDU (i.e., data analysis, human reliability analysis, systems analysis, etc.). Therefore, the attributes described in this

6

publication apply as well for these reactor types. There is however one area regarding PSA approaches and techniques where there is a significant difference. The concept of core damage as an accident sequence end state for Level-1 PSA and as a rough measure for consequences is a useful concept for vessel type LWRs. The physical background for this concept is that for the compact cores of current vessel type LWRs once there is loss of cooling to substantial portions of the core and associated fuel damage it is likely that the whole core is affected and a substantial part of the fission product inventory is released from the fuel.

For reactors with physically well-separated fuel channels and comparatively large

cores, damage might be restricted to individual fuel channels, small portions of the core, parts of the core or may extend to the entire core. Accordingly, several Level-1 fuel or core damage end-states have been defined and used in PSAs for such reactors to reflect the significantly different fractions of the fuel or core affected during different accident scenarios. The physical reason for this distinction and refinement of core damage are specific features of the reactor and system design, e.g. the design of coolant piping and the connection of emergency core cooling system (ECCS) trains to the coolant piping. The refined definition of core damage then allows obtaining a useful consequence measure in terms of the Level-1 PSA by distinguishing scenarios with significant consequences from those with low consequences but elevated frequencies. In turn such definition of core damage categories requires interpretation and adaptation regarding the description of PSA tasks and associated attributes as given in this publication.

It should be pointed out however that analytical and in particular experimental

information on details of the accident progression in the beyond design accident range is limited for other reactor types if compared to vessel type LWRs. This may also affect the formulation of safety function success criteria and delineation of accident sequences. Therefore, accident progression and the characterization of Level-1 end states for reactor types other than vessel type LWRs should be regarded as being still under development.

1.6.2. Applicability for NPPs in different stages of the plant’s lifetime

In order to provide a comprehensive description of PSA tasks and associated attributes the publication is focused on PSA approaches, modelling aspects, and data for a typical ‘mature’ nuclear power plant, which typically has been in operation for a number of years without major changes in the plant. It is recognized that significant differences exist regarding PSA approaches, modelling aspects, and data for different stages of the plants lifetime. The reason why this publication concentrates on PSA for a ‘mature’ nuclear power plant is that only at this stage the full range of PSA techniques can be applied including evaluation and use of a reasonable amount of operational experience data from the plant itself. During the design stage of a plant, for example, detailed information on design and operational features might be limited and no operational experience data from the plant is available. For a completely new design even applicable experience data from comparable plants might not be available. A number of attributes formulated in this publication for a ‘mature’ NPP therefore require interpretation and adaptation when applied for an NPP in an earlier life stage.

PSA techniques can be used beginning at an early design stage of a plant. At this time

even the conceptual design of engineered safety features (ESFs) might not be entirely fixed, e.g. the number of redundant trains in an ECCS. Diverse ECCSs might be under consideration for a particular emergency core cooling function. In these situations, PSA techniques can be used to support conceptual decisions. However, there are major differences in PSA approaches, techniques and data as listed below:

7

• Detailed design information on systems and their support systems might not or only partially be available. This missing information can be bridged by related assumptions for PSA purposes.

• Detailed operating procedures are not available. Information from similar NPPs might be used instead.

• No or only generic operational experience is available.

• Information on thermal hydraulic analyses, accident progression, accident scenarios might be limited.

• Completeness of initiating events is difficult to ascertain.

• Limited information on man-machine interface (MMI) and on training of operating staff is available.

• Limited information on maintenance practices and procedures.

• Equipment location information is limited or missing (important for CCF modelling and modelling of secondary effects).

• Details on technical specifications (TSs) are missing or limited.

In summary, a considerable number of attributes which apply for a PSA for a ‘mature’ plant do not strictly apply for a plant in the design stage, simply because the required knowledge and data are not yet available. Accordingly, simplifications and assumptions need to be made to bridge the missing information. Furthermore, the applicability and adequacy of the assumptions themselves could be limited, introducing elements of variability and uncertainty even if not directly visible or stated. A typical example for a new NPP concept in an early design stage are the advanced reactors incorporating new concepts for safety features, e.g. passive systems, where even the assessment of thermal-hydraulics and consideration of reliability aspects are still under development and where operational experience is not available.

Uncertainty inherent in PSA models and results for a reactor in the design stage needs

to be fully addressed if the PSA is used for decision making. One way to deal with these uncertainties in the early stages of plant life is to use PSA in a relative way, e.g. by comparing different design variants using similar models and assumptions. Another useful approach is to check the robustness of results by changing the assumptions to see how the results are affected by such changes.

During the further development of the plant, increasing details on design and

operational features of the plant become available. This progress is then usually reflected in refining PSA models, which in turn reduces associated variabilities and uncertainties.

Even for a ‘mature’ plant, there might be major backfits or major changes regarding

the plants design and operational features. A major change in this sense would be a significant redesign of the core, including redesign of protection instrumentation and control (I&C) and of ESFs. This in turn would mean that the PSA would need to be redone because such a major change is likely to change the entire PSA model structure and because similar conditions would apply again for parts of the plant or for the entire plant as during the design stage, which has to be reflected in the PSA techniques and data.

8

2. OVERVIEW OF PSA APPLICATIONS 2.1. PSA application categorization

Since the beginning of the nineties, PSA techniques have been used increasingly widely in many countries in the risk-informed decision making process in NPP design, operation, and licensing activities. IAEA-TECDOC-1200 (see Ref. [7]), published in 2001, identified a number of PSA applications. In this publication, some additional applications have been identified, and the PSA applications have been categorized in the following way according to their purpose:

1. Safety assessment: to assess the overall safety of the plant and to develop an understanding of the main contributors to risk

2. Design evaluation: to provide support for design evaluation

3. NPP operation: to provide support for day-to-day operation of the plant (not including permanent changes to design or operational practices)

4. Permanent changes to the operating plant: to assess the safety significance of proposed permanent changes to the plant design, hardware, or administrative controls (e.g. operating procedures, the licensing basis) as an aid to decision making

5. Oversight activities: to support plant performance monitoring and assessment (both regulatory and industry)

6. Evaluation of safety issues: to evaluate safety issues Several application groups can be defined under the six categories based on a more

specific consideration of the purpose and subject of PSA applications. Table 2.1 provides a list of PSA application categories, groups within the categories, and specific applications within the groups.

2.2. PSA results and metrics used in decision making

In order to use a PSA in the decision making process, it is necessary to define what

results are needed, and define criteria these results may be compared with. In some cases, the results may be qualitative, but in most cases, the results are quantitative. In such cases, some parameters that can be calculated using the PSA model are defined, which are referred to as metrics. Typically, used metrics are importance measures, core damage frequency (CDF), large early release frequency (LERF), conditional core damage probability (CCDP), quantitative health objectives (QHO), etc. The metrics most commonly used are defined in Appendix I. Although analysis of the LERF and the QHO typically are not in the scope of a Level-1 PSA (and this publication), these metrics can be derived and used in the decision making process in many PSA applications and therefore for the sake of completeness are also addressed in Appendixes I and II of this publication. It should be noted that some PSA applications require a wider PSA scope, e.g. emergency planning would require in principle a full-scope Level-3 PSA. The PSA scope and metrics used in decision making provide an input to the definition of application-specific PSA quality requirements. In the decision making process the evaluated metrics are compared against some decision criteria that need to be established. The various forms that such decision criteria can take are discussed elsewhere (e.g. see Ref. [7]).

9

Table 2.1 PSA Applications

Application Category Application Group Specific Application

1.1 Assessment of the overall plant safety

1.2 Periodic safety review

1. SAFETY ASSESSMENT

1.3 Analysis of the degree of defence against assumed terrorist attack scenarios

2.1 Application of PSA to support decisions made during the NPP design (plant under design)

2. DESIGN EVALUATION

2.2 Assessment of the safety importance of deviations between an existing plant design and updated/revised deterministic design rules

3.1.1 Maintenance program optimization 3.1.2 Risk-informed house keeping

3.1 NPP maintenance

3.1.3 Risk-informed support for plant ageing management program

3.2.1 Development and improvement of the emergency operating procedures

3.2.2 Support for NPP accident management (severe accident prevention, severe accident mitigation)

3.2 Accident mitigation and emergency planning

3.2.3 Support for NPP emergency planning 3.3.1 Improvement of operator training program 3.3.2 Improvement of maintenance personnel training

program

3.3 Personnel training

3.3.3 Improvement of plant management training program 3.4.1 Configuration planning (e.g. support for plant

maintenance and test activities) 3.4.2 Real time configuration assessment and control

(response to emerging conditions) 3.4.3 Exemptions to TS and justification for continued

operation

3. NPP OPERATION

3.4 Risk-based configuration control/ Risk Monitors

3.4.4 Dynamic risk-informed TS

4.1.1 NPP upgrades, back-fitting activities and plant modifications

4.1 Plant changes

4.1.2 Lifetime extension

4.2.1 Determination and evaluation of changes to allowed outage time and changes to required TS actions

4.2.2 Risk-informed optimisation of TS 4.2.3 Determination and evaluation of changes to

surveillance test intervals 4.2.4 Risk-informed in-service testing (IST)

4.2 Technical specification changes

4.2.5 Risk-informed in-service inspections (RI-ISI)

4.3.1 Equipment risk significance evaluation

4. PERMANENT CHANGES TO THE OPERATING PLANT

4.3 Establishment of graded QA program for SSC

4.3.2 Evaluation of risk impact of changes to QA requirements

10

Application Category Application Group Specific Application

5.1.1 Planning and prioritization of inspection activities (regulatory and industry)

5.1.2 Long term risk-based performance indicators

5.1 Performance monitoring

5.1.3 Short term risk based performance indicators 5.2.1 Assessment of inspection findings

5. OVERSIGHT ACTIVITIES

5.2 Performance assessment 5.2.2 Evaluation and rating of operational events

6.1.1 Risk evaluation of corrective measures 6.1 Risk evaluation 6.1.2 Risk evaluation to identify and rank safety issues 6.2.1 Long term regulatory decisions

6. EVALUATION OF SAFETY ISSUES

6.2 Regulatory decisions 6.2.2 Interim regulatory decisions

11

3. PROCEDURE TO ACHIEVE QUALITY IN PSA APPLICATIONS This section is devoted to the description of the general approach for the presentation

of information in the publication including definitions of PSA elements and attributes, the coding scheme for naming general and special PSA attributes, and how the publication can be applied for the purpose of assessing and enhancing the PSA quality. 3.1. PSA elements and attributes

The PSA features (termed ‘attributes’ in this publication) are provided for the nine

PSA elements that comprise an internal events, at-power, Level-1 PSA. The PSA elements identify the major analysis areas. It should be noted that while the nine PSA elements are identified, this division is to some extent arbitrary because all the analysis areas are interconnected and influence each other.

The PSA elements and associated abbreviations used in this publication are the

following:

1. Initiating Events Analysis IE (Section 4)

2. Accident Sequence Analysis AS (Section 5)

3. Success Criteria Formulation and Supporting Analysis SC (Section 6)

4. Systems Analysis SY (Section 7)

5. Human Reliability Analysis HR (Section 8)

6. Data Analysis DA (Section 9)

7. Dependent Failures Analysis DF (Section 10)

8. Model Integration and Core Damage Frequency Quantification MQ (Section 11)

9. Results Analysis and Interpretation RI (Section 12) Each PSA element is described in a separate section of the publication as indicated

above including a description of the objectives of the analysis relating to the PSA element, a list of major analysis tasks, and tables describing general and special attributes for the tasks.

The general and special attributes are defined as follows:

• General attributes describe the main features of the analyses, documentation, and data to be considered in the ‘base case PSA’. These features are described based on the state of the art, as defined by generally accepted good practice, of a PSA constructed to evaluate the core damage frequency.

• Special attributes describe information on specific features of PSA elements to be

satisfied in order that the PSA could be considered as appropriate for a specific application.

It is assumed that the general attributes characterise a contemporary state of the art

Level-1 internal events at-power PSA performed with the aim to assess the overall NPP safety. Special attributes provide elevated capabilities within particular PSA elements to meet special features of PSA applications. Sections 4-12 present general and special attributes for

12

the nine PSA elements listed above. An identifier is assigned to each general and special attribute in accordance with the coding scheme provided in Section 3.2. Several special attributes may be defined for a general attribute. They are provided together: special attributes underneath of the corresponding general attribute. The tables include the identifiers of general attributes in the first column (‘GA’ in the table heading), the description of general attributes and, where appropriate, the identifiers/descriptions of the associated special attributes (in Italics) in the second column, as well as rationale/comments/examples for general attributes and special attributes (in Italics) in the third column.

If a PSA meets solely the general attributes, it would not necessarily mean that the

PSA could be used consistently and reliably for any PSA application, e.g. the applications listed in Section 2. Sometimes, special attributes may be important to enhance the depth and level of detail of the analysis in specific areas to facilitate the use of PSA for specific applications. Section 13 discusses what special attributes are appropriate for particular applications and how to determine them.

It should be noted that the PSA results used in the decision making process might be

adequate even if certain attributes are not met or not met fully. However, this would generally require that either a demonstration that the attribute is not required to produce the results needed to support the application, or that the decision has compensated for this failure to meet the attribute, by, for example, restricting the scope of the application to that supported by the PSA results. The methods by which this may be demonstrated are not within the scope of the publication.

3.2. Coding scheme for attributes identifiers (1) General attribute The identifier of a general attribute is represented by the following string:

XX-YNN,

where: XX – the identifier of a PSA element as provided in Section 3.1 (IE, DA, etc.); Y – a letter (in alphabetic order) designating the task within the PSA element; NN – a two-digit number designating the sequential number of the general attribute within

Task ‘Y’.

Example: IE-A01 - this is the identifier of the first general attribute for Task ‘A’ of the PSA element ‘IE’.

(2) Special attribute

The identifier of a special attribute represents the following string: XX-YNN-SM,

where: XX-YNN – the identifier of the general attribute, for which a special attribute is provided; S – the letter ‘S’ indicating that a special attribute is defined for General Attribute

‘XX-YNN’;

13

M – an one-digit number designating the sequential number of the special attribute relating to the considered general attribute.

Example: IE-A01-S1 - this is the identifier of the first special attribute related to the first

general attribute for Task ‘A’ of the PSA element ‘IE’.

3.3. Connection to IAEA PSA guidelines The PSA elements #1 through 9 applicable to a Level-1 at-power internal event PSA

are principally addressed in the IAEA Level-1 PSA Procedure Guide (see Ref. [2]). In order to provide a link between that publication and the present publication, a mapping of the PSA tasks from Ref. [2] to the PSA elements considered in this publication is provided in Appendix III.

There are also several IAEA publications providing more details on the following PSA

elements:

• Human Reliability Analysis (Ref. [6]) • Initiating Events Analysis (Ref. [11]) • Dependent Failures Analysis (Ref. [5]).

3.4. The procedure for use of the TECDOC

This publication has two major uses:

(1) Support the PSA review as carried out by the IAEA within IPSART missions and as part of other activities.

(2) Support the use of PSA applications in planning of a PSA project, to make sure that

appropriate quality of the Level-1 internal events PSA is achieved, or assessing the applicability of an existing PSA intended for use in an application.

The general procedure of application of the approach provided in this publication for

assessing and enhancing the PSA quality involves consideration of general and special attributes. The following three major steps are involved:

STEP (1) For the PSA application, a consideration needs to be given to plant design and

operational features affected by the application (and related PSA models). PSA scope and results/metrics required for the application have to be determined. In case the scope and results/metrics are insufficient, there may be a need to refine, complete, or upgrade the PSA. For instance, in case an application (e.g. severe accident management) requires Level-2 PSA results, but these are not available, an extension of the PSA scope would be needed. Alternatively, the developer could provide supplementary arguments/analysis/data to bridge limitations, or restrict the scope of the application to that supported by the PSA.

STEP (2) For each PSA element, a determination needs to be made whether the general

attributes provided in Sections 4-12 characterizing a ‘base case PSA’ have been met. If not, refinements should be considered for the PSA to bring it in compliance with the general attributes. Alternatively, the developer could provide a supplementary justification to demonstrate that the general attributes not met are not required for the application.

14

STEP (3) For the PSA application, a determination needs to be made whether the special attributes needed for this PSA application have been met. Section 13 of this publication should be consulted regarding the practical steps for identification of a set of special attributes of the PSA elements relevant for the application of interest. A determination needs to be made whether the special attributes have been made. If not, refinements should be considered for the PSA to bring it in compliance with the special attributes.

The procedure for determination of quality of PSA for applications is illustrated in

Figure 2. For Steps 2 and 3 depicted in the figure, this publication covers only a Level-1 internal events at-power PSA. In case a PSA includes more PSA modules (e.g. internal floods, fires, shutdown PSA, etc.), other publications should be consulted regarding the technical features of the analyses, e.g. Ref. [8] for internal floods, Ref. [12] for external hazards.

15

The PSA results can support the application

Fig. 2. General procedure for determination of quality of PSA for applications.

Refine, complete, or upgrade PSA

YES

OR

STEP 1 Are the PSA scope and results/metrics

sufficient?

Identify type of PSA application (e.g. Table 2.1)

Identify plant design and operational features affected by the application

Identify PSA scope and results/metrics required (desired) for the application

Refine PSA elements acc. to Special Attributes

Refine PSA elements acc. to General Attributes

Provide supplementary arguments/analysis/data to

bridge limitations, OR restrict the scope of the application to

that supported by the PSA

NO

YES

YES

NO

NO

Demonstrate/justify that the attributes not met are not

required for the application OR

STEP 2 Do PSA elements satisfy General

Attributes?

STEP 3 Do PSA elements

relevant for the application satisfy Special

Attributes?

16

4. PSA ELEMENT ‘IE’: INITIATING EVENTS ANALYSIS

4.1. Main objectives

The initiating events analysis is a highly iterative, multi-purpose task, which provides the basis for the PSA and ensures its completeness. The risk profile can be incomplete and distorted if important initiating events (IEs) are omitted or incorrectly included in the IE groups.

The main objectives of the initiating events analysis are as follows:

• to identify a reasonably complete set of the events that interrupt normal plant operation and that require successful mitigation to prevent core damage, so that no significant contributor to core damage is omitted;

• to group initiating events to facilitate the efficient modelling of plant response and

initiating events frequency assessment while providing sufficient resolution regarding modelling of accident sequences (events included in the same group have similar mitigation requirements, or are bounded by the limiting mitigation requirements for the ‘representative initiating event’ for the group);

• to provide estimates for the frequencies of the initiating event groups using information

available and associated estimation techniques. Important aspects of the IE analysis are the following:

• Initiating event definition is correct and complete.

• Initiating events are identified taking into account all plant configurations possible at power operation.

• IEs are grouped in a consistent manner such that any event in the group has the same or less demanding mitigation requirements than initiating event chosen as the IE group representative for further modelling.

• Methods used for the estimation of the IE frequencies are clearly distinguished for the cases when:

- Estimation is based on plant specific or generic, or both kinds of statistical information.

- Estimation is based on system models (mainly refers to system initiators and includes checking system failures which may create an initiating event).

- Estimation for rare events, which is based on expert judgment or use of specific methods (e.g. structural mechanics analysis, etc.)

• Uncertainties in the IE frequencies are understood, evaluated, accounted for, and documented.

17

4.2. Initiating events analysis tasks and their attributes

The main tasks for the PSA element ‘IE Analysis’ are listed in Table 4.2. Tables 4.2-A through 4.2-H present the description of general and special attributes for these tasks.

Table 4.2 Main Tasks for IE Analysis

Task ID Task Content

IE-A Identification of IE Candidates (Preliminary Identification of IEs)

IE-B IE Screening and Final IEs List Identification

IE-C IEs Grouping

IE-D Collection and Evaluation of Generic Information for IE Frequencies Assessment

IE-E Collection of Plant-Specific Information

IE-F Handling the Effects of Plant Modifications

IE-G IE Frequencies Quantification

IE-H Documentation

18

Tab

le 4

.2-A

A

ttri

bute

s for

IE A

naly

sis:

Tas

k IE

-A ‘I

dent

ifica

tion

of IE

Can

dida

tes’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Sp

ecia

l Attr

ibut

es (i

n Ita

lics)

IE-A

A

list

of i

nitia

ting

even

ts is

def

ined

whi

ch is

as c

ompl

ete

as p

ossi

ble.

IE-A

01

IE d

efin

ition

is c

lear

and

cov

ers a

ny p

lant

dis

turb

ance

s tha

t req

uire

miti

gatio

n to

pr

even

t cor

e da

mag

e.

CO

MM

ENT :

The

follo

win

g IE

def

initi

on is

gen

eral

ly u

sed:

“IE

is a

n ev

ent w

hich

co

uld

dire

ctly

lead

to c

ore

dam

age

or c

halle

nges

nor

mal

pla

nt o

pera

tion

and

requ

ires s

ucce

ssfu

l miti

gatio

n to

pre

vent

cor

e da

mag

e”.

The

initi

atin

g ev

ents

id

entif

ied

typi

cally

incl

ude

trans

ient

s of v

ario

us ty

pes,

Loss

of C

oola

nt A

ccid

ents

(L

OC

As)

, int

erfa

cing

sys

tem

s LO

CA

s, st

eam

gen

erat

or tu

be ru

ptur

es, a

nd su

ppor

t sy

stem

initi

ator

s.

IE-A

02

All

plan

t ope

ratio

ns m

odes

with

pow

er o

pera

tion

are

cons

ider

ed to

geth

er w

ith th

eir

inte

rlock

s and

syst

em c

onfig

urat

ion.

IEs a

pplic

able

to sp

ecifi

c co

nfig

urat

ions

are

id

entif

ied.

RA

TIO

NA

LE: S

yste

ms c

onfig

urat

ion,

inte

rlock

s and

requ

irem

ents

for p

lant

sh

utdo

wn

may

be

diff

eren

t for

the

sam

e pl

ant o

pera

ting

on d

iffer

ent p

ower

leve

ls.

This

may

lead

to th

e ap

pear

ance

of a

dditi

onal

IEs a

nd/o

r to

chan

ges i

n th

e bo

unda

ry c

ondi

tions

of t

he IE

s, w

hich

wer

e id

entif

ied

for n

omin

al p

ower

leve

l.

EXA

MPL

ES:

At s

ome

plan

ts (e

.g. V

VER

-440

NPP

s) th

e fo

llow

ing

diff

eren

ces r

elat

ed to

pla

nt

pow

er o

pera

tion

mod

e ex

ist:

1) E

vent

invo

lvin

g bo

ron

dilu

tion

due

to e

rron

eous

con

nect

ion

of a

dis

conn

ecte

d lo

op is

pos

sibl

e on

ly a

t the

leve

l low

er th

an n

omin

al p

ower

.

2) T

rip o

f 3 M

CPs

lead

s to

reac

tor s

cram

onl

y at

pow

er le

vel b

elow

75%

from

no

min

al.

3) T

rip o

f las

t ope

ratin

g tu

rbin

e le

ads t

o im

med

iate

reac

tor s

cram

onl

y w

hen

the

plan

t ope

rate

s at p

ower

leve

l bel

ow 7

5% fr

om n

omin

al.

4) P

lant

con

figur

atio

n (in

par

ticul

ar se

cond

ary

side

arr

ange

men

ts) d

iffer

s si

gnifi

cant

ly w

hile

the

unit

oper

ates

at 5

0% a

nd 1

00%

pow

er le

vel (

one

and

two

turb

ines

).

IE-A

03

A st

ruct

ured

, sys

tem

atic

pro

cess

for i

dent

ifyin

g in

itiat

ing

even

ts is

em

ploy

ed. T

he

follo

win

g m

etho

ds fo

r ide

ntifi

catio

n of

pot

entia

l IEs

are

use

d:

1) A

naly

sis o

f lis

ts o

f IEs

from

PSA

s for

sim

ilar u

nits

:

List

s of I

Es d

evel

oped

in P

SAs f

or si

mila

r uni

ts a

re re

view

ed in

ord

er to

id

entif

y po

tent

ial I

Es a

pplic

able

to th

e in

vest

igat

ed u

nit.

2) A

naly

sis o

f gen

eric

list

s of I

Es:

The

lists

of I

Es fr

om g

ener

ic so

urce

s (e.

g. g

ener

ic li

sts f

rom

IAEA

, US

NR

C,

EPR

I, et

c.) f

or si

mila

r uni

ts a

re re

view

ed in

ord

er to

iden

tify

pote

ntia

l IEs

RA

TIO

NA

LE: U

se o

f onl

y on

e or

a su

bset

of t

he m

etho

ds li

sted

in IE

-A03

may

no

t pro

vide

a c

ompl

ete

list o

f ini

tiatin

g ev

ents

due

to in

here

nt li

mita

tions

of e

ach

met

hod.

CO

MM

ENT:

For

spec

ific

appl

icat

ions

the

use

of a

subs

et o

f the

met

hods

may

pr

ovid

e a

list o

f IEs

suff

icie

nt to

dea

l with

the

appl

icat

ion.

Thi

s sho

uld

incl

ude

at

leas

t: -

Pre

viou

s PSA

s lis

ts fo

r sim

ilar u

nits

. -

Ope

ratio

nal e

xper

ienc

e of

the

unit

unde

r con

side

ratio

n an

d si

mila

r uni

ts.

19

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

Spec

ial A

ttrib

utes

(in

Italic

s)

appl

icab

le to

the

inve

stig

ated

uni

t.

3) A

naly

sis o

f ope

ratio

nal e

xper

ienc

e:

- O

pera

tiona

l exp

erie

nce

of th

e un

it un

der c

onsi

dera

tion

and

sim

ilar u

nits

is

revi

ewed

in o

rder

to id

entif

y ev

ents

that

hap

pene

d in

the

past

.

- Th

e ex

perie

nce

on p

lant

-spe

cific

initi

atin

g ev

ents

is re

view

ed to

ass

ure

that

th

e lis

t of c

halle

nges

acc

ount

s for

the

plan

t exp

erie

nce.

Exp

erie

nce

and

anal

yses

at s

imila

r pla

nts a

re re

view

ed to

ass

ess w

heth

er th

e lis

t of c

halle

nges

in

clud

ed in

the

mod

el a

ccou

nts f

or th

e in

dust

ry e

xper

ienc

e.

4) D

educ

tive

anal

ysis

(e.g

. mas

ter l

ogic

dia

gram

, hea

t bal

ance

faul

t tre

es, e

tc.)

A st

ep-b

y st

ep c

onse

quen

tial a

naly

sis i

s per

form

ed in

ord

er to

iden

tify

even

ts,

whi

ch c

ould

lead

to c

ore

dam

age

or re

quire

miti

gatio

n ac

tions

.

5) In

duct

ive

anal

ysis

(e.g

. fai

lure

mod

e ef

fect

ana

lysi

s [FM

EA])

.

Syst

emat

ic e

valu

atio

n of

eac

h sy

stem

is p

erfo

rmed

to a

sses

s the

pos

sibi

lity

of

an in

itiat

ing

even

t occ

urrin

g du

e to

a fa

ilure

of t

he s

yste

m, e

.g. d

etai

led

mod

el

of sy

stem

inte

rfac

es in

clud

ing

faul

t tre

e de

velo

pmen

t and

/or F

MEA

to a

sses

s an

d do

cum

ent t

he p

ossi

bilit

y of

an

initi

atin

g ev

ent r

esul

ting

from

indi

vidu

al

syst

ems/

syst

em tr

ain/

equi

pmen

t fai

lure

s. A

ll sy

stem

s, w

hich

failu

res m

ay

brin

g di

stur

banc

e in

pla

nt o

pera

tion

(e.g

. nor

mal

ope

ratio

n, fr

ont-l

ine

and

supp

ort s

yste

ms)

are

revi

ewed

, with

the

exce

ptio

n of

thos

e, w

hich

wer

e al

read

y id

entif

ied

as th

e so

urce

of I

Es b

ased

on

othe

r ana

lysi

s. T

he a

naly

ses

are

perf

orm

ed a

fter s

yste

m m

odel

s wer

e de

velo

ped.

6) L

ists

of d

esig

n ba

sis a

ccid

ents

(DB

As)

and

bey

ond

desi

gn b

asis

acc

iden

ts

(BD

BA

s) a

re re

view

ed.

IE-A

04

Initi

atin

g ev

ents

resu

lting

from

mul

tiple

failu

res a

nd re

quiri

ng d

iffer

ent m

itiga

tion

stra

tegy

are

incl

uded

, in

parti

cula

r tho

se th

at c

an re

sult

from

a c

omm

on c

ause

. R

ATI

ON

ALE

: Eve

nts c

ause

d by

mul

tiple

equ

ipm

ent f

ailu

res m

ay b

e si

gnifi

cant

in

term

s of r

isk

even

if th

e IE

freq

uenc

y is

low

.

See

also

Tab

le 1

0.2-

D, G

ener

al A

ttrib

ute

DF-

D01

.

IE-A

05

The

plan

t ope

ratio

ns, m

aint

enan

ce, e

ngin

eerin

g, a

nd sa

fety

ana

lysi

s per

sonn

el a

re

inte

rvie

wed

to d

eter

min

e if

any

pote

ntia

l ini

tiatin

g ev

ents

hav

e be

en o

verlo

oked

. In

terv

iew

info

rmat

ion

from

sim

ilar p

lant

s is a

lso

used

.

RA

TIO

NA

LE: I

nter

view

of e

xper

ienc

ed p

lant

per

sona

l may

giv

e ad

ditio

nal

know

ledg

e on

real

pla

nt b

ehav

iour

and

may

hel

p to

iden

tify

IEs o

verlo

oked

with

th

e us

e of

met

hods

list

ed in

IE-A

03.

20

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

Spec

ial A

ttrib

utes

(in

Italic

s)

IE-A

06

Syst

em o

r equ

ipm

ent f

ailu

res o

r com

bina

tions

of t

hese

are

scre

ened

to se

e w

heth

er

they

repr

esen

t eith

er a

uni

que

IE, w

hich

nee

ds se

para

te tr

eatm

ent o

r rep

rese

nt a

co

ntrib

utin

g IE

to a

n IE

gro

up.

Dep

ende

ncie

s bet

wee

n sy

stem

initi

ator

s and

pos

t-tri

p fu

nctio

ns n

eed

to b

e id

entif

ied

in th

is p

roce

ss. (

See

also

Tab

le 1

0.2-

D, T

ask

DF-

D).

All

syst

em in

itiat

ors t

hat a

t the

sam

e tim

e re

sult

in p

ost-t

rip fu

nctio

n fa

ilure

s are

id

entif

ied.

CO

MM

ENT:

The

se ty

pes o

f eve

nts a

re k

now

n as

com

mon

cau

se in

itiat

ors o

r sy

stem

initi

ator

s. A

com

mon

cau

se in

itiat

or (C

CI)

is a

n ev

ent c

ausi

ng a

tran

sien

t (o

r req

uirin

g m

anua

l shu

tdow

n) a

nd a

t the

sam

e tim

e de

grad

ing

one

or m

ore

safe

ty

func

tions

that

may

be

need

ed a

fter t

he tr

ansi

ent/s

hut-d

own.

EXA

MPL

ES:

Failu

re m

odes

, whi

ch d

isru

pt n

orm

al o

pera

tion

are:

-

Nor

mal

ly o

pera

ting

pum

ps: f

ailu

re to

run

- St

andb

y pu

mps

: ina

dver

tent

star

t-up/

failu

re to

star

t -

Safe

ty/re

lief v

alve

s: in

adve

rtent

ope

ning

-

Nor

mal

ly c

lose

d bu

s bre

aker

: ina

dver

tent

ope

ning

Typi

cally

, the

con

trol a

nd p

rote

ctio

ns sy

stem

s, el

ectri

c po

wer

supp

ly sy

stem

, se

rvic

e w

ater

syst

em o

r oth

er su

ppor

t sys

tem

s are

sour

ces f

or u

nexp

ecte

d C

CIs

, w

here

the

plan

t's tr

ansi

ent e

xper

ienc

e ca

nnot

pro

vide

info

rmat

ion.

The

follo

win

g ar

e th

e m

ain

area

s for

iden

tific

atio

n of

CC

Is:

Loss

of p

roce

ss c

ontro

l. A

n an

alys

is o

f the

pro

cess

con

trol i

nclu

des b

oth

mea

sure

men

t and

con

trol o

f pro

cess

par

amet

ers.

A la

rge

num

ber o

f par

amet

ers

exis

t in

the

plan

t, w

hich

are

use

d to

supe

rvis

e an

d co

ntro

l the

pla

nt, p

ower

, lev

el,

pres

sure

, flo

w, t

empe

ratu

re, h

umid

ity, e

tc. L

oss o

f som

e pa

ram

eter

s may

cau

se (o

r re

quire

) a p

lant

trip

and

func

tiona

lly d

egra

de o

ne o

r mor

e sa

fety

sys

tem

s, e.

g.:

- Er

rone

ous l

evel

mea

sure

men

t in

the

reac

tor v

esse

l -

Spur

ious

isol

atio

n si

gnal

s.

Loss

of p

ower

supp

ly. S

ome

failu

res i

n th

e po

wer

supp

ly w

hich

may

cau

se (o

r re

quire

) a p

lant

trip

and

func

tiona

lly d

egra

de o

ne o

r mor

e sa

fety

sys

tem

s, e.

g.:

- Lo

ss o

f ext

erna

l pow

er

- Lo

ss o

f spe

cific

alte

rnat

e cu

rren

t (A

C) o

r dire

ct c

urre

nt (D

C) b

usba

rs.

Loss

of a

uxili

ary

syst

ems.

Som

e fa

ilure

s with

in a

uxili

ary

syst

ems m

ay c

ause

(or

requ

ire) a

pla

nt tr

ip a

nd fu

nctio

nally

deg

rade

one

or m

ore

safe

ty re

late

d sy

stem

s, e.

g.:

- Lo

ss o

f ins

trum

ent a

ir -

Loss

of c

oolin

g w

ater

Com

pone

nt fa

ilure

s in

safe

ty sy

stem

s. C

ompo

nent

failu

res i

n sa

fety

syst

ems m

ay

degr

ade

safe

ty fu

nctio

ns, a

nd m

ay a

lso

be a

requ

irem

ent f

or a

pla

nt sh

ut-d

own

(Tec

hnic

al S

peci

ficat

ion

rule

s).

21

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

Spec

ial A

ttrib

utes

(in

Italic

s)

Sp

ecia

l Attr

ibut

e IE

-A06

-S1:

Ev

ents

cau

sed

by o

pera

tor e

rror

s are

iden

tifie

d an

d fu

rthe

r ev

alua

ted.

RATI

ON

ALE:

Usu

ally

eve

nts c

ause

d by

ope

rato

r err

ors a

re a

ssum

ed to

be

take

n in

to c

onsi

dera

tion

in IE

freq

uenc

ies e

stim

ated

on

the

basi

s of o

pera

tiona

l ex

peri

ence

. H

owev

er, f

or c

erta

in a

pplic

atio

ns im

plic

it co

nsid

erat

ion

of IE

s ca

used

by

oper

ator

err

ors m

ay h

ide

the

impa

ct o

f cha

nges

in p

lant

mai

nten

ance

an

d op

erat

iona

l pra

ctic

e. T

his i

s par

ticul

arly

impo

rtan

t for

the

com

mon

cau

se

initi

ator

s bec

ause

of t

he p

oten

tial d

epen

denc

y be

twee

n th

e er

ror a

ssoc

iate

d w

ith

the

initi

atin

g ev

ent a

nd th

ose

actio

ns re

quir

ed to

resp

ond

to th

e ev

ent.

IE-A

07

Even

ts c

ause

d by

equ

ipm

ent d

amag

e du

ring

on p

ower

refu

ellin

g pr

oces

s are

co

nsid

ered

as p

oten

tial I

Es.

CO

MM

ENT:

App

licab

le fo

r the

pla

nts w

ith a

t-pow

er re

fuel

ling

(i.e.

CA

ND

U,

RB

MK

, AG

R, M

AG

NO

X, v

esse

l typ

e he

avy

wat

er re

acto

r [A

ttuch

a]).

IE-A

08

Initi

atin

g ev

ent p

recu

rsor

s and

run-

back

eve

nts (

also

cal

led

hous

e-lo

ad tu

rbin

e op

erat

ion)

are

revi

ewed

for t

he p

urpo

se o

f IE

iden

tific

atio

n.

RA

TIO

NA

LE:

Rev

iew

of I

E pr

ecur

sors

and

run-

back

eve

nts m

ay h

elp

to id

entif

y IE

s ove

rlook

ed w

ith th

e us

e of

met

hods

list

ed in

IE-A

03 a

nd p

rovi

des a

par

tial

basi

s for

qua

ntify

ing

thei

r fre

quen

cies

CO

MM

ENTS

:

1) A

run-

back

eve

nt is

an

even

t, w

hich

doe

s not

resu

lt in

a p

lant

trip

if p

lant

run-

back

syst

ems a

re su

cces

sful

.

2) A

pre

curs

or fo

r an

IE is

a k

ind

of tr

igge

r eve

nt, w

hich

alo

ne d

oes n

ot re

pres

ent

an in

itiat

ing

even

t, bu

t tog

ethe

r with

oth

er e

vent

s may

cau

se a

n IE

. A sp

ecia

l an

alys

is m

ay b

e ne

cess

ary

to m

odel

thes

e ev

ents

in th

e PS

A IE

s, e.

g. a

n IE

eve

nt

tree.

EXA

MPL

E: T

urbi

ne tr

ips,

mai

n co

olan

t pum

p tri

p, m

ain

feed

wat

er p

ump

failu

re

etc.

hav

e to

be

cons

ider

ed ta

king

into

acc

ount

the

avai

labi

lity

of a

utom

atic

ca

pabi

litie

s, e.

g. fo

r pow

er re

duct

ion,

to a

void

a re

acto

r scr

am.

IE-A

09

Even

ts th

at h

ave

occu

rred

at c

ondi

tions

oth

er th

an a

t pow

er o

pera

tion

(i.e.

dur

ing

shut

dow

n co

nditi

ons)

are

iden

tifie

d an

d ex

amin

ed o

n th

e ap

plic

abili

ty fo

r at p

ower

op

erat

ion.

RA

TIO

NA

LE: T

he IE

s hav

e oc

curr

ed a

t shu

tdow

n co

nditi

ons m

ay h

ave

a po

tent

ial t

o oc

cur a

t pow

er o

pera

tion.

The

se e

vent

s sho

uld

be in

clud

ed in

the

list o

f IE

s, un

less

it is

just

ified

that

the

IE c

anno

t phy

sica

lly o

ccur

at p

ower

ope

ratio

n.

How

ever

, the

use

of t

hese

eve

nts i

n da

ta tr

eatm

ent s

houl

d be

mad

e w

ith c

are.

22

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

Spec

ial A

ttrib

utes

(in

Italic

s)

IE-A

10

Adm

inis

trativ

e (o

rder

ly) s

hutd

own

caus

ed b

y di

ffere

nt re

ason

s (e.

g. fa

ilure

of

sing

le o

r mul

tiple

trai

ns o

f fro

nt-li

ne o

r sup

port

syst

ems)

is in

clud

ed in

the

list o

f IE

s. Th

ese

IEs a

re re

view

ed in

ord

er to

avo

id d

oubl

e co

untin

g in

shut

dow

n PS

A.

RA

TIO

NA

LE: A

dmin

istra

tive

shut

dow

n fo

r cer

tain

reas

ons m

ay c

ause

risk

si

gnifi

cant

acc

iden

t seq

uenc

es. E

xclu

sion

of I

Es o

f thi

s typ

e m

ay h

inde

r cer

tain

ap

plic

atio

ns (e

.g. t

hose

one

s dea

ling

with

exe

mpt

ions

to T

S, ju

stifi

catio

n fo

r co

ntin

ued

oper

atio

n, e

tc.)

EXA

MPL

E: A

dmin

istra

tive

orde

rly sh

utdo

wn

due

to re

quire

men

ts o

f TS

(e.g

. ex

ceed

ing

the

AO

T af

ter a

failu

re o

f the

em

erge

ncy

feed

wat

er p

ump

iden

tifie

d at

pe

riodi

cal t

est).

IE-A

11

Mul

ti-un

it si

te in

itiat

ors a

re id

entif

ied

and

incl

uded

in th

e lis

t of I

Es a

s ‘M

ulti-

units

in

itiat

ors’

.

RA

TIO

NA

LE: A

n IE

may

be

caus

ed b

y sy

stem

/equ

ipm

ent f

ailu

res a

t ano

ther

uni

t at

mul

tiple

-uni

t site

. The

se IE

s cou

ld n

ot b

e id

entif

ied

with

the

use

of m

etho

ds

liste

d in

IE-A

03; h

owev

er th

ey m

ay b

e si

gnifi

cant

con

tribu

tors

to th

e ris

k in

pa

rticu

lar t

o th

ose

mul

tiple

-uni

ts p

lant

s with

shar

ed e

quip

men

t.

EXA

MPL

E: F

or th

e m

ulti-

units

pla

nt w

hen

two

or m

ore

units

shar

e th

e sa

me

build

ing

(e.g

. tur

bine

hal

l) IE

s occ

urre

d at

one

uni

t may

aff

ect n

orm

al o

pera

tion

of

the

othe

r uni

t.

CO

MM

ENT:

Mul

ti-un

its in

itiat

ors m

ay in

clud

e th

e ev

ents

occ

urre

d bo

th a

t the

ne

arby

uni

ts a

nd a

t the

uni

t und

er c

onsi

dera

tion.

The

mai

n fe

atur

e of

thes

e in

itiat

ors t

hat t

hey

affe

ct se

vera

l uni

ts e

ither

by

caus

ing

the

dist

urba

nce

on th

e ne

arby

uni

t or b

y sh

arin

g co

mm

on e

quip

men

t. Th

e la

st fe

atur

e is

impo

rtant

for I

E gr

oupi

ng ta

sk (S

ee T

able

4.2

-C, G

ener

al A

ttrib

ute

IE-C

08).

IE-A

12

The

even

ts d

ealin

g w

ith fa

ilure

s of i

ndiv

idua

l sup

port

syst

ems (

or tr

ains

) tha

t can

ca

use

a pl

ant t

rip a

re in

clud

ed in

the

list o

f IEs

. C

OM

MEN

T: F

ailu

res o

f sup

port

syst

em tr

ain(

s) th

at m

ay c

ause

an

initi

atin

g ev

ent

shou

ld b

e in

clud

ed in

the

IE li

st in

add

ition

to th

e fa

ilure

of w

hole

syst

em.

If o

nly

the

failu

re o

f who

le su

ppor

t sys

tem

is in

clud

ed in

the

IE li

st a

s a si

ngle

mos

t co

nser

vativ

e ev

ent,

the

freq

uenc

y of

such

eve

nts m

ay b

e un

dere

stim

ated

.

IE-A

13

The

list o

f pot

entia

l ini

tiatin

g ev

ents

is d

evel

oped

bas

ed o

n th

e re

sults

of t

he

anal

yses

with

in th

e ta

sk IE

-A.

23

T

able

4.2

-B

Att

ribu

tes f

or IE

Ana

lysi

s: T

ask

IE-B

‘IE

Scr

eeni

ng a

nd F

inal

IEs L

ist I

dent

ifica

tion’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Sp

ecia

l Attr

ibut

es (i

n Ita

lics)

IE-B

A

ll id

entif

ied

even

ts a

re su

bjec

ted

to a

scre

enin

g an

alys

is in

ord

er to

scre

en o

ut e

vent

s not

ap

plic

able

for t

he u

nit u

nder

con

side

ratio

n an

d to

com

pile

a fi

nal l

ist o

f IEs

.

IE-B

01

The

even

ts a

re sc

reen

ed o

ut fr

om th

e lis

t of I

Es a

nd e

limin

ated

from

furth

er c

onsi

dera

tion

only

if c

ompl

ianc

e w

ith o

ne o

f the

follo

win

g cr

iteria

is ju

stifi

ed:

a) T

he e

vent

doe

s not

cor

resp

ond

to th

e ac

cept

ed IE

def

initi

on.

b) T

he e

vent

doe

s not

cor

resp

ond

to th

e sc

ope

of th

e PS

A.

c) T

he fr

eque

ncy

of th

e ev

ent i

s les

s tha

n th

e tru

ncat

ion

valu

e re

late

d to

the

freq

uenc

y of

a

sign

ifica

nt a

ccid

ent s

eque

nce,

and

the

even

t doe

s not

invo

lve

eith

er a

n IS

LOC

A,

cont

ainm

ent b

ypas

s, or

reac

tor p

ress

ure

vess

el ru

ptur

e. F

or th

ese

even

ts th

e tru

ncat

ion

valu

e is

at l

east

one

ord

er o

f mag

nitu

de lo

wer

than

the

trunc

atio

n va

lue

acce

pted

in th

e PS

A.

d) T

he re

sulti

ng re

acto

r shu

tdow

n is

not

an

imm

edia

te o

ccur

renc

e. T

hat i

s, th

e ev

ent d

oes

not r

equi

re th

e pl

ant t

o tra

nsfe

r to

shut

dow

n co

nditi

ons u

ntil

suff

icie

nt ti

me

has e

xpire

d du

ring

whi

ch th

e in

itiat

ing

even

t con

ditio

ns, w

ith a

hig

h de

gree

of c

erta

inty

(bas

ed o

n su

ppor

ting

calc

ulat

ions

), ar

e de

tect

ed a

nd c

orre

cted

bef

ore

norm

al p

lant

ope

ratio

n is

cu

rtaile

d (e

ither

adm

inis

trativ

ely

or a

utom

atic

ally

).

RA

TIO

NA

LE: T

he e

vent

s sho

uld

not b

e sc

reen

ed o

ut if

a p

oten

tial f

or a

hi

gh L

evel

-2 c

ontri

butio

n is

reco

gniz

ed.

Sp

ecia

l Attr

ibut

e IE

-B01

-S1:

Ev

ents

with

freq

uenc

y be

low

the

trun

catio

n va

lue

are

revi

site

d in

the

scre

enin

g pr

oces

s and

reta

ined

in th

e lis

t of I

Es to

iden

tify

a w

ider

ra

nge

of p

ossi

ble

haza

rds.

EXAM

PLE:

Cha

nges

in p

lant

test

and

mai

nten

ance

pra

ctic

e m

ay le

ad to

an

incr

ease

of t

he fr

eque

ncie

s of t

hese

IEs.

Excl

usio

n of

IEs o

f thi

s typ

e m

ay m

ask

thei

r pot

entia

l im

port

ance

for c

erta

in a

pplic

atio

ns.

IE-B

02

IEs,

the

freq

uenc

ies o

f whi

ch sh

ould

be

asse

ssed

by

FT m

odel

ling,

are

iden

tifie

d (e

.g.

supp

ort s

yste

m fa

ilure

s, C

CI,

etc.

). R

ATI

ON

ALE

: Fau

lt tre

e m

odel

ling

allo

ws t

o pr

oper

ly ta

ke in

to a

ccou

nt

supp

ort o

r aux

iliar

y sy

stem

dep

ende

ncie

s and

com

preh

ensi

vely

acc

ount

fo

r diff

eren

t cau

sal m

echa

nism

s to

estim

ate

IEs f

requ

enci

es.

IE-B

03

The

final

list

of I

Es is

dev

elop

ed. A

ll ev

ents

, whi

ch w

ere

not s

cree

ned

out a

re in

clud

ed in

th

e fin

al li

st.

24

T

able

4.2

-C

Att

ribu

tes f

or IE

Ana

lysi

s: T

ask

IE-C

‘IE

Gro

upin

g’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

IE-C

IE

s are

gro

uped

in se

para

te g

roup

s with

sim

ilar m

itiga

tion

requ

irem

ents

for a

ll IE

s in

the

grou

p in

ord

er to

faci

litat

e an

eff

icie

nt, b

ut re

alis

tic e

stim

atio

n of

CD

F (e

.g. m

anag

eabl

e nu

mbe

r of a

ccid

ent s

eque

nce

mod

el a

nd su

ffic

ient

info

rmat

ion

for f

requ

ency

est

imat

ion)

.

The

diffe

rent

IE g

roup

s are

cha

ract

eriz

ed b

y di

ffer

ent i

mpa

cts o

n pl

ant p

erfo

rman

ce, s

afet

y fu

nctio

ns, a

nd p

ossi

bilit

ies f

or re

cove

ry.

CO

MM

ENT:

IEs a

re g

roup

ed in

sepa

rate

gro

up if

it is

impo

rtant

for

spec

ific

appl

icat

ion

IE-C

01

A st

ruct

ured

, sys

tem

atic

pro

cess

for g

roup

ing

the

initi

atin

g ev

ents

is u

sed.

The

acc

iden

t pr

ogre

ssio

ns a

nd su

cces

s crit

eria

are

iden

tifie

d fo

r eac

h of

the

IE (a

vaila

ble

ther

mal

hy

drau

lic a

naly

sis a

nd e

xper

t jud

gmen

t is u

sed)

. (Se

e al

so S

ectio

ns 5

and

6).

RA

TIO

NA

LE: G

roup

ing

can

be p

erfo

rmed

onl

y ba

sed

on si

mila

rity

of

acci

dent

pro

gres

sion

and

succ

ess c

riter

ia o

f all

even

t inc

lude

d in

the

grou

p.

IE-C

02

Initi

atin

g ev

ents

are

gro

uped

in a

sing

le g

roup

onl

y w

hen

the

follo

win

g ca

n be

ass

ured

:

- Ev

ents

hav

e th

e sa

me

safe

and

uns

afe

end

stat

es a

nd le

ad to

sim

ilar a

ccid

ent

prog

ress

ion

in te

rms o

f pla

nt re

spon

se, s

ucce

ss c

riter

ia, t

imin

g, a

nd th

e ef

fect

on

the

oper

abili

ty o

f rel

evan

t miti

gatin

g sy

stem

s and

ope

rato

rs p

erfo

rman

ce; o

r

- Ev

ents

can

be

subs

umed

into

a g

roup

and

bou

nded

by

the

wor

st c

ase

impa

cts w

ithin

th

e ‘n

ew’ g

roup

.

RA

TIO

NA

LE: M

eetin

g of

the

requ

ired

cond

ition

s ens

ures

that

any

sp

ecifi

c fe

atur

e of

the

IE in

clud

ed in

the

grou

p w

as n

ot tr

eate

d in

op

timis

tic m

anne

r and

ther

efor

e no

pot

entia

l ins

ight

s of t

he P

SA a

re

over

look

ed.

CO

MM

ENT:

Tho

se IE

s tha

t hav

e si

gnifi

cant

ly d

iffer

ent e

nviro

nmen

tal

impa

ct o

r cou

ld h

ave

mor

e se

vere

radi

onuc

lide

rele

ase

pote

ntia

l are

gr

oupe

d se

para

tely

from

oth

er in

itiat

ing

even

t cat

egor

ies.

EXA

MPL

E: S

uch

initi

ator

s as i

nter

faci

ng sy

stem

s LO

CA

, SG

tube

ru

ptur

es, h

igh-

ener

gy st

eam

line

bre

aks o

utsi

de c

onta

inm

ent a

re u

sual

ly

mod

elle

d as

a se

para

te g

roup

s.

Sp

ecia

l Attr

ibut

e IE

-C02

-S1:

IE

with

a re

lativ

ely

low

freq

uenc

y, b

ut m

ore

seve

re a

ccid

ent

prog

ress

ion

and

mor

e de

man

ding

succ

ess c

rite

ria

(com

pari

ng to

the

othe

r IEs

in th

e gr

oup)

is in

clud

ed in

a se

para

te IE

gro

up.

RATI

ON

ALE:

For

cer

tain

app

licat

ions

a re

alis

tic re

pres

enta

tion

of p

lant

re

spon

se to

war

ds se

vere

IEs w

ith lo

w fr

eque

ncy

is re

quir

ed.

Sp

ecia

l Attr

ibut

e IE

-C02

-S2:

If

the

sign

als f

or a

ctua

tion

of a

t lea

st o

ne m

itiga

tion

syst

em a

re d

iffer

ent

for d

iffer

ent I

Es, t

hese

IEs a

re in

clud

ed in

sepa

rate

gro

ups.

RATI

ON

ALE :

For

cer

tain

app

licat

ions

a re

alis

tic re

pres

enta

tion

of th

e pl

ant r

espo

nse

tow

ards

IEs w

ith d

iffer

ent s

igna

ls fo

r sys

tem

s act

uatio

n an

d op

erat

ion

is im

port

ant.

EXAM

PLE:

SLO

CA

and

Pres

suri

zer s

team

leak

may

hav

e di

ffere

nt

sign

als f

or a

ctua

tion

of H

PEC

C p

umps

(e.g

. low

pre

ssur

e an

d lo

w le

vel i

n pr

essu

rize

r for

SLO

CA

and

only

low

pre

ssur

e in

pre

ssur

ize

for

pres

suri

zer l

eak)

.

25

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

IE-C

03

IEs a

re in

clud

ed in

sepa

rate

gro

ups i

f diff

eren

t ope

rato

r act

ions

or c

ondi

tions

for o

pera

tor

actio

ns in

term

s of i

nfor

mat

ion

avai

labl

e, ti

me

win

dow

s, en

viro

nmen

tal c

ondi

tions

, and

pr

oced

ural

requ

irem

ents

exi

st.

EXA

MPL

E: E

vent

s req

uirin

g IS

LOC

A is

olat

ion

shou

ld b

e co

nsid

ered

se

para

tely

for e

ach

ISLO

CA

pat

h if

diff

eren

t iso

latio

n po

ssib

ilitie

s are

av

aila

ble.

IE-C

04

Plan

t spe

cific

ther

mal

hyd

raul

ic a

naly

ses s

uppo

rting

acc

iden

t seq

uenc

e m

odel

ling

and

succ

ess c

riter

ia d

efin

ition

for t

he re

pres

enta

tive

IE in

eac

h gr

oup

are

perf

orm

ed.

In c

ase

the

succ

ess c

riter

ia fo

r fre

quen

t eve

nts i

nclu

ded

in th

e gr

oup

are

muc

h le

ss se

vere

th

an su

cces

s crit

eria

for t

he re

pres

enta

tive

IE, t

he e

vent

s sho

uld

be g

roup

ed d

iffer

ently

. (S

ee a

lso

IE-C

02-S

1.)

RA

TIO

NA

LE: T

his a

ttrib

ute

help

s to

avoi

d ex

cess

ive

cons

erva

tism

.

IE-C

05

Each

com

mon

cau

se in

itiat

or is

con

side

red

as a

sepa

rate

gro

up.

RA

TIO

NA

LE: G

roup

ing

of th

is ty

pe o

f eve

nts c

ould

mas

k th

e in

sigh

ts

for c

erta

in a

pplic

atio

ns.

IE-C

06

‘Mul

tiple

-uni

ts in

itiat

ors’

aff

ectin

g sy

stem

s/eq

uipm

ent s

hare

d be

twee

n se

vera

l uni

ts a

re

cons

ider

ed a

s sep

arat

e gr

oups

. R

ATI

ON

ALE

: The

cap

acity

of m

itiga

tion

syst

ems m

ay b

e si

gnifi

cant

ly

diffe

rent

whe

n m

ore

than

one

uni

t is a

ffect

ed b

y th

e IE

.

EXA

MPL

E: C

omm

on c

oold

own

syst

em a

t VV

ER-4

40/2

30 c

ould

not

pr

ovid

e ef

ficie

nt c

oolin

g if

both

uni

ts a

re p

ut in

to sh

utdo

wn

cond

ition

s si

mul

tane

ousl

y.

IE-C

07

The

IEs,

for w

hich

the

stra

tegy

of a

ccid

ent m

itiga

tion

depe

nds o

n th

e pl

ace

of th

eir

orig

inat

ion

(e.g

. diff

eren

t pos

sibi

litie

s for

leak

isol

atio

n or

diff

eren

t im

pact

on

oper

atio

n of

ot

her e

quip

men

t), a

re c

onsi

dere

d in

sepa

rate

gro

ups u

nles

s the

impa

ct is

bou

nded

by

the

wor

st-c

ase

loca

tion.

EXA

MPL

ES:

1) F

or so

me

plan

ts, w

here

LO

CA

isol

atio

n is

pos

sibl

e, L

OC

A in

isol

able

an

d no

n-is

olab

le p

arts

can

be

cons

ider

ed.

2) F

or st

eam

lines

bre

aks o

utsi

de a

nd in

side

con

tain

men

t, th

e en

viro

nmen

tal c

ondi

tions

impo

rtant

for o

pera

tion

of o

ther

equ

ipm

ent m

ay

be si

gnifi

cant

ly d

iffer

ent.

IE-C

08

A li

st o

f IE

grou

ps is

com

pile

d. A

repr

esen

tativ

e IE

for f

urth

er m

odel

ling

of e

ach

IE g

roup

is

sele

cted

. The

stric

test

feat

ures

in te

rms o

f acc

iden

t pro

gres

sion

and

succ

ess c

riter

ia o

f an

IE in

clud

ed in

the

grou

p ar

e as

sign

ed fo

r the

repr

esen

tativ

e IE

.

CO

MM

ENT:

The

hyp

othe

tical

IE th

at c

ombi

nes t

he w

orst

suc

cess

crit

eria

of

all

IEs i

n th

e gr

oup

may

be

cons

truct

ed fo

r fur

ther

ana

lysi

s.

26

T

able

4.2

-D

Att

ribu

tes f

or IE

Ana

lysi

s: T

ask

IE-D

‘Col

lect

ing

and

Eva

luat

ing

Gen

eric

Info

rmat

ion

for

IE F

requ

enci

es A

sses

smen

t’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

IE-D

G

ener

ic in

form

atio

n re

quire

d to

per

form

the

IE fr

eque

ncie

s ass

essm

ent i

s col

lect

ed.

Gen

eric

info

rmat

ion

is e

valu

ated

rega

rdin

g its

app

licab

ility

. App

licab

le g

ener

ic so

urce

s are

se

lect

ed fo

r eac

h IE

/IE g

roup

.

CO

MM

ENT:

Gen

eric

info

rmat

ion

is n

eede

d w

hen

plan

t spe

cific

dat

a is

no

t ava

ilabl

e, a

s wel

l as f

or th

e pu

rpos

e of

Bay

esia

n up

datin

g.

IE-D

01

Gen

eric

info

rmat

ion

requ

ired

for t

he IE

freq

uenc

y es

timat

ion

is c

olle

cted

in o

rder

to

acco

unt f

or a

bro

ader

exp

erie

nce.

Bas

ical

ly th

e fo

llow

ing

gene

ric in

form

atio

n da

ta is

requ

ired:

a)

Num

ber o

f IEs

ver

sus p

lant

ope

ratio

nal t

ime.

b)

Freq

uenc

ies o

f IE

for r

are

even

ts.

c)

Des

crip

tion

of th

e m

etho

ds u

sed

and

cond

ition

s und

er w

hich

gen

eric

info

rmat

ion

was

ob

tain

ed.

d)

Uni

t typ

e w

here

dat

a ca

me

from

(i.e

. pla

nt d

esig

n)

e)

IE d

efin

ition

and

bou

ndar

y co

nditi

ons i

n ge

neric

sour

ces

f)

Uni

t ope

ratin

g m

ode

to w

hich

IE re

cord

ed is

rela

ted

The

sour

ces o

f gen

eric

dat

a an

d th

e pr

oces

s of t

he d

eriv

atio

n of

gen

eric

IE fr

eque

ncie

s es

timat

es a

re id

entif

ied

and

desc

ribed

. The

IE d

efin

ition

s and

bou

ndar

y co

nditi

ons a

re

eval

uate

d in

the

view

of c

onsi

sten

cy w

ith th

e ev

ents

det

erm

ined

in th

e ta

sk IE

-B.

Gen

eric

info

rmat

ion

is tr

aced

to th

e pr

imar

y so

urce

to a

void

inad

verte

nt d

oubl

e co

untin

g of

in

form

atio

n an

d to

ass

ure

that

no

info

rmat

ion

from

the

plan

t is c

onta

ined

in th

e ge

neric

in

form

atio

n. T

his i

s im

porta

nt fr

om th

e vi

ewpo

int o

f con

sist

ent B

ayes

ian

upda

ting.

(See

al

so T

able

4.2

-H, T

ask

IE-H

).

RA

TIO

NA

LE: A

pplic

abili

ty o

f dat

a fr

om p

lant

s of d

iffer

ent d

esig

n sh

ould

be

inve

stig

ated

and

bou

ndar

ies o

f the

IEs i

n th

e ge

neric

sour

ce

shou

ld b

e co

mpa

red

with

pla

nt-s

peci

fic IE

s bou

ndar

ies

IE-D

02

The

colle

ctio

n an

d ev

alua

tion

of g

ener

ic in

form

atio

n in

clud

e an

und

erst

andi

ng a

nd

asse

ssm

ent o

f the

app

licab

ility

and

the

unce

rtain

ty in

the

orig

inal

dat

a.

RA

TIO

NA

LE: I

nfor

mat

ion

on th

e un

certa

inty

of g

ener

ic d

ata

supp

orts

th

e de

cisi

on o

n th

e ap

plic

abili

ty o

f the

gen

eric

dat

a fo

r Bay

esia

n up

datin

g w

ith p

lant

spec

ific

data

.

IE-D

03

The

gene

ric in

form

atio

n co

llect

ed is

eva

luat

ed in

ord

er to

iden

tify

the

info

rmat

ion

appl

icab

le fo

r a sp

ecifi

c IE

and

/or I

E gr

oup.

R

ATI

ON

ALE

: In

case

app

licab

ility

of g

ener

ic d

ata

to th

e pl

ant s

peci

fic

IEs i

s not

just

ified

, spe

cial

con

side

ratio

n sh

ould

be

give

n on

the

poss

ibili

ty to

app

ly B

ayes

ian

upda

ting

proc

ess.

27

T

able

4.2

-E

Att

ribu

tes f

or IE

Ana

lysi

s: T

ask

IE-E

‘Col

lect

ing

of P

lant

-Spe

cific

Info

rmat

ion’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

IE-E

Pl

ant-s

peci

fic in

form

atio

n re

quire

d to

per

form

IE fr

eque

ncie

s ass

essm

ent i

s col

lect

ed in

ac

cord

ance

with

the

IEs l

ists

des

crib

ed in

the

attri

bute

s of t

he ta

sk IE

-B a

nd IE

gro

upin

g ou

tline

d in

the

attri

bute

s of I

E-C

.

IE-E

01

Plan

t-spe

cific

ope

ratio

nal i

nfor

mat

ion

in a

ccor

danc

e w

ith th

e IE

s lis

ts re

fere

nced

in th

e at

tribu

tes o

f the

task

s IE-

B a

nd IE

-C is

col

lect

ed.

RA

TIO

NA

LE: P

lant

spec

ific

even

ts a

re c

olle

cted

in o

rder

to p

rovi

de

info

rmat

ion,

whi

ch is

pla

nt sp

ecifi

c an

d re

flect

s pla

nt d

esig

n an

d op

erat

iona

l fea

ture

s.

IE-E

02

The

data

base

con

tain

ing

info

rmat

ion

on e

vent

s and

pla

nt o

pera

tiona

l his

tory

is c

reat

ed, w

ith

the

poss

ibili

ty to

ext

ract

the

follo

win

g da

ta:

- nu

mbe

r of I

Es fo

r eac

h IE

gro

up;

- nu

mbe

r of p

recu

rsor

s for

the

IEs;

- nu

mbe

r of r

un-b

ack

even

ts;

- un

it op

erat

ion

mod

e;

- du

ratio

n of

IE (e

.g. f

or o

ffsi

te p

ower

);

- tim

e pe

riod

of d

ata

colle

ctio

n;

- de

scrip

tion

of th

e ev

ent.

CO

MM

ENT:

Dur

atio

n of

an

IE m

ay b

e im

porta

nt fo

r est

imat

ion

of

freq

uenc

ies o

f the

IEs o

f cer

tain

dur

atio

n (i.

e. L

OO

P IE

).

IE-E

03

Whe

n sy

stem

mod

els f

or in

itiat

ing

even

ts fr

eque

ncy

asse

ssm

ent a

re u

sed,

the

appr

opria

te

syst

em in

form

atio

n is

col

lect

ed. (

See

also

Tab

le 7

.2-C

, Gen

eral

Attr

ibut

e SY

-C08

, and

Ta

ble

10.2

-D, T

ask

DF-

D).

28

T

able

4.2

-F

Att

ribu

tes f

or IE

Ana

lysi

s: T

ask

IE-F

‘Han

dlin

g of

Eff

ects

of P

lant

Mod

ifica

tions

’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

IE-F

D

iffer

ence

s bet

wee

n hi

stor

ical

pla

nt a

vaila

bilit

y ov

er th

e pe

riod

of e

vent

occ

urre

nces

in th

e pl

ant d

atab

ase

and

pres

ent p

lant

ava

ilabi

lity,

whi

ch c

ould

be

diffe

rent

from

his

toric

al

valu

es, a

re a

ccou

nted

for.

IE-F

01

Col

lect

ed in

form

atio

n on

pla

nt-s

peci

fic IE

s is c

heck

ed fo

r whe

ther

the

caus

es fo

r the

IEs

are

appl

icab

le fo

r the

act

ual p

lant

con

ditio

ns.

RA

TIO

NA

LE: I

f the

pla

nt c

hang

es m

ake

the

IEs t

hat h

appe

ned

in th

e pa

st

inap

plic

able

to th

e ac

tual

pla

nt s

tate

then

the

anal

ysis

of t

he im

pact

of t

he

chan

ges

and

asse

ssm

ent o

f the

hyp

othe

tical

eff

ect o

n th

e hi

stor

ical

dat

a to

de

term

ine

to w

hat e

xten

t the

dat

a ca

n be

use

d sh

ould

be

perf

orm

ed.

IE-F

02

Plan

t mod

ifica

tions

/cha

nges

are

eva

luat

ed in

ord

er to

iden

tify

whe

ther

the

prec

urso

rs a

nd

run-

back

eve

nts h

appe

ned

in th

e pa

st m

ay c

ause

an

IE fo

r the

act

ual p

lant

con

ditio

ns.

RA

TIO

NA

LE: S

ee IE

-F01

.

IE-F

03

The

ratio

nale

for s

cree

ning

or d

isre

gard

ing

plan

t-spe

cific

dat

a is

just

ified

(e.g

. pla

nt d

esig

n m

odifi

catio

ns, c

hang

es in

ope

ratin

g pr

actic

es).

RA

TIO

NA

LE: S

ee IE

-F01

.

IE-F

04

The

excl

usio

n of

ear

lier y

ears

that

are

not

repr

esen

tativ

e of

cur

rent

dat

a is

just

ified

. R

ATI

ON

ALE

: Thi

s attr

ibut

e he

lps t

o av

oid

exce

ssiv

e co

nser

vatis

m.

Act

ual e

xper

ienc

e sh

ows t

hat d

urin

g fir

st y

ears

of p

lant

ope

ratio

n a

num

ber o

f IEs

sign

ifica

ntly

hig

her t

han

afte

r cer

tain

per

iod.

Sp

ecia

l Attr

ibut

e IE

-F04

-S1:

Tim

e tr

end

anal

ysis

is u

sed

to a

ccou

nt fo

r est

ablis

hed

tren

ds, e

.g.

decr

easi

ng re

acto

r tri

p ra

tes i

n re

cent

yea

rs.

RATI

ON

ALE :

Tim

e tr

end

anal

ysis

may

be

impo

rtan

t for

cer

tain

ap

plic

atio

ns.

EXAM

PLE:

Ris

k ev

alua

tion

of th

e m

easu

res i

mpl

emen

ted

with

the

aim

to

elim

inat

e sp

ecifi

c IE

s.

IE-F

05

An

anal

ysis

aim

ed to

iden

tify

whe

ther

new

IEs a

re in

trodu

ced

by p

lant

m

odifi

catio

ns/c

hang

es is

per

form

ed.

EXA

MPL

E: R

epla

cem

ent o

f ana

logu

e by

dig

ital I

&C

may

lead

to

intro

duci

ng o

f new

IEs.

29

T

able

4.2

-G

Att

ribu

tes f

or IE

Ana

lysi

s: T

ask

IE-G

‘IE

Fre

quen

cies

Qua

ntifi

catio

n’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

IE-G

Fr

eque

ncy

of in

divi

dual

IE a

nd/o

r IE

grou

p is

ass

esse

d ba

sed

on re

leva

nt g

ener

ic in

dust

ry

and

plan

t-spe

cific

evi

denc

e. W

here

feas

ible

, gen

eric

and

pla

nt-s

peci

fic e

vide

nce

is

inte

grat

ed u

sing

acc

epta

ble

met

hods

to o

btai

n pl

ant-s

peci

fic IE

freq

uenc

y es

timat

es. I

E fr

eque

ncy

estim

atio

n is

acc

ompa

nied

by

a ch

arac

teriz

atio

n of

the

unce

rtain

ty.

IE-G

01

Initi

atin

g ev

ent f

requ

enci

es a

re c

alcu

late

d ta

king

into

acc

ount

the

frac

tion

of ti

me

the

plan

t is a

t pow

er.

RA

TIO

NA

LE: U

se o

f cal

enda

r yea

rs fo

r fre

quen

cy e

stim

atio

n fo

r the

PSA

at

pow

er o

pera

tion

lead

s to

unde

rest

imat

ion

of th

e IE

freq

uenc

ies.

IE-G

02

Rea

listic

IE fr

eque

ncy

estim

ates

are

cal

cula

ted

usin

g B

ayes

ian

upda

tes w

here

feas

ible

. Pr

ior d

istri

butio

ns a

re se

lect

ed a

s eith

er n

on-in

form

ativ

e, o

r rep

rese

ntat

ive

of v

aria

bilit

y in

indu

stry

dat

a.

IE-G

03

For r

are

initi

atin

g ev

ents

, the

indu

stry

gen

eric

dat

a is

use

d w

ith a

ccou

nt fo

r pla

nt-

spec

ifics

. R

ATI

ON

ALE

: Use

of B

ayes

ian

upda

ting

with

zer

o pl

ant-s

peci

fic st

atis

tics

shou

ld b

e pe

rfor

med

with

car

e in

ord

er to

avo

id d

oubl

e co

untin

g of

the

expo

sure

tim

e po

tent

ially

acc

ount

ed in

indu

stry

gen

eric

dat

a.

CO

MM

ENT:

‘Rar

e ev

ent’

is a

n ev

ent t

hat m

ight

be

expe

cted

to o

ccur

onc

e or

a fe

w ti

mes

thro

ugho

ut th

e w

orld

nuc

lear

indu

stry

exp

erie

nce.

IE-G

04

For e

xtre

mel

y ra

re in

itiat

ing

even

ts a

n en

gine

erin

g ju

dgm

ent i

s use

d, a

ugm

ente

d w

ith

appl

icab

le g

ener

ic d

ata

sour

ces a

nd sp

ecifi

c an

alys

is (e

.g. s

truct

ural

mec

hani

cs m

etho

ds,

etc.

).

RA

TIO

NA

LE: U

se o

f any

stat

istic

al a

naly

sis m

etho

ds (i

.e. t

he B

ayes

ian

upda

ting)

cou

ld n

ot p

rodu

ce u

sefu

l res

ults

for t

he ‘e

xtre

mel

y ra

re’ e

vent

s du

e to

pra

ctic

al a

bsen

ts o

f sta

tistic

al in

form

atio

n.

CO

MM

ENT:

‘Ext

rem

ely

rare

eve

nt’ i

s an

even

t tha

t wou

ld n

ot b

e ex

pect

ed

to o

ccur

eve

n on

ce th

roug

hout

the

indu

stry

exp

erie

nce.

IE-G

05

Gen

eric

and

pla

nt sp

ecifi

c da

ta a

re u

sed

in a

just

ifiab

le m

anne

r. W

hen

the

Bay

esia

n ap

proa

ch is

use

d to

der

ive

a di

strib

utio

n an

d m

ean

valu

e of

IE fr

eque

ncy,

the

chec

k is

m

ade

that

the

post

erio

r dis

tribu

tions

der

ived

are

reas

onab

le g

iven

the

prio

r dis

tribu

tions

an

d th

e pl

ant s

peci

fic e

vide

nce.

CO

MM

ENT:

If th

e es

timat

or fo

r the

mea

n va

lue

of a

par

amet

er b

ased

on

plan

t evi

denc

e is

out

side

of a

95%

con

fiden

ce in

terv

al a

roun

d th

e m

edia

n va

lue

of th

e pr

ior d

istri

butio

n th

e ap

plic

abili

ty o

f tha

t par

ticul

ar p

rior d

ata

and

dist

ribut

ion

shou

ld b

e re

cons

ider

ed re

gard

ing

its a

pplic

abili

ty to

the

initi

atin

g ev

ent u

nder

con

side

ratio

n.

30

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

IE-G

06

Whi

le u

sing

syst

em m

odel

s for

initi

atin

g ev

ents

freq

uenc

y as

sess

men

t the

syst

em m

odel

s ar

e m

odifi

ed in

such

a w

ay th

at (S

ee a

lso

Tabl

e 4.

2-B

, Gen

eral

Attr

ibut

e IE

-B02

):

- IE

freq

uenc

ies a

re q

uant

ified

(rat

her t

han

prob

abili

ties)

.

- A

ll re

leva

nt c

ombi

natio

ns o

f eve

nts i

nvol

ving

the

annu

al fr

eque

ncy

of o

ne

com

pone

nt fa

ilure

com

bine

d w

ith th

e un

avai

labi

lity

(or f

ailu

re d

urin

g th

e re

pair

time

of th

e fir

st c

ompo

nent

) of o

ther

com

pone

nts a

re c

aptu

red.

- Pr

obab

ility

of s

purio

us a

ctua

tion

of th

e eq

uipm

ent a

nd C

CF

are

cons

ider

ed a

nd

acco

unte

d fo

r.

- Sp

urio

us a

ctua

tion

of th

e eq

uipm

ent a

nd p

assi

ve fa

ilure

s not

con

side

red

in th

e PS

A

mod

el a

re in

clud

ed in

the

mod

el.

- Su

ppor

t sys

tem

failu

res a

re in

clud

ed in

the

mod

el (u

nles

s the

supp

ort s

yste

m is

an

own

initi

ator

).

- Pr

e-ac

cide

nt h

uman

err

ors a

re in

clud

ed in

the

mod

el.

- Im

pact

of n

earb

y un

its is

acc

ount

ed fo

r.

RA

TIO

NA

LE: T

his a

ttrib

ute

help

s to

avoi

d un

dere

stim

atio

n of

IE

freq

uenc

ies f

or th

e IE

s tre

ated

with

the

use

of F

T m

odel

ling

tech

niqu

e.

CO

MM

ENT:

It i

s im

porta

nt to

cor

rect

ly m

odel

in F

Ts th

e fo

llow

ing

aspe

cts o

f sys

tem

ope

ratio

n:

- po

ssib

ilitie

s for

reco

very

of t

he re

dund

ant e

quip

men

t; -

mis

sion

tim

e co

rres

pond

ing

to a

ctua

l sys

tem

(com

pone

nt)

oper

atin

g tim

e w

ithin

a y

ear,

incl

udin

g C

CF

(e.g

. 800

0 h

inst

ead

of

24 h

use

d in

the

acc

iden

t seq

uenc

es a

naly

ses)

, etc

.

IE-G

07

Each

syst

em a

lignm

ent a

nd a

lignm

ents

of s

uppo

rting

syst

ems t

hat c

ould

influ

ence

the

likel

ihoo

d th

at fa

ilure

s cau

se a

n in

itiat

ing

even

t, or

mag

nify

the

seve

rity

of th

e ch

alle

nge

to p

lant

safe

ty fu

nctio

ns th

at w

ould

resu

lt fr

om su

ch a

n ev

ent i

s acc

ount

ed fo

r.

RA

TIO

NA

LE: R

ealig

nmen

t of t

he e

quip

men

t may

lead

to in

crea

se o

f the

fr

eque

ncy

of s

yste

m fa

ilure

s cau

sing

the

IE.

EXA

MPL

E: A

t som

e pl

ants

dur

ing

test

of r

eact

or p

rote

ctio

n sy

stem

one

sy

stem

trai

n is

out

of o

pera

tion.

Dur

ing

this

per

iod

the

syst

em c

onfig

urat

ion

(2 o

ut o

f 3) c

hang

es (1

out

of 2

), w

hich

may

lead

to a

sign

ifica

nt in

crea

se in

th

e fr

eque

ncy

of sp

urio

us a

ctua

tion

of th

e re

acto

r scr

am.

IE-G

08

In th

e IS

LOC

A fr

eque

ncy

anal

ysis

, tho

se fe

atur

es o

f pla

nt a

nd p

roce

dure

s tha

t cou

ld

sign

ifica

ntly

influ

ence

the

ISLO

CA

freq

uenc

y ar

e ac

coun

ted

for.

EXA

MPL

E: A

bsen

ce o

f tes

t pos

sibi

lity

for o

ne c

heck

val

ve in

the

sequ

ence

of

two

lead

to si

gnifi

cant

incr

ease

of t

he e

vent

with

leak

age

due

to h

igh

prob

abili

ty th

at th

e se

cond

che

ck v

alve

s is i

n th

e fa

iled

stat

e.

IE-G

09

The

freq

uenc

y of

IEs,

for w

hich

the

stra

tegy

of a

ccid

ent m

itiga

tion

depe

nds o

n th

e pl

ace

of th

eir o

rigin

atio

n (e

.g. d

iffer

ent p

ossi

bilit

ies f

or le

ak is

olat

ion

or d

iffer

ent i

mpa

ct o

n op

erat

ion

of o

ther

equ

ipm

ent),

is e

stim

ated

for t

he sp

ecifi

c lo

catio

n ta

king

into

acc

ount

ge

omet

rical

cha

ract

eris

tics o

f the

NPP

s pip

elin

es.

EXA

MPL

ES:

1) F

or so

me

plan

ts, w

here

SLO

CA

isol

atio

n is

pos

sibl

e, S

LOC

A in

isol

able

an

d no

n-is

olab

le p

art c

an b

e ca

lcul

ated

by

parti

tioni

ng o

f the

tota

l fre

quen

cy

taki

ng in

to a

ccou

nt th

e le

ngth

of t

he p

ipel

ines

bef

ore

and

afte

r the

isol

atio

n va

lve.

2) F

or st

eam

line

s bre

aks,

the

freq

uenc

y of

the

IEs o

ccur

red

insi

de a

nd

outs

ide

cont

ainm

ent c

an b

e ca

lcul

ated

by

parti

tioni

ng o

f the

tota

l IE

freq

uenc

y ta

king

into

acc

ount

the

leng

th o

f the

pip

elin

es in

side

and

out

side

co

ntai

nmen

t.

31

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

IE-G

10

Plan

t-spe

cific

info

rmat

ion

is u

sed

in th

e as

sess

men

t and

qua

ntifi

catio

n of

reco

very

act

ions

w

here

ava

ilabl

e fo

r IE

freq

uenc

ies e

stim

atio

n. T

hese

reco

very

act

ions

shou

ld b

e cl

early

de

fined

in o

rder

to a

void

dou

ble

cred

iting

in IE

freq

uenc

y as

sess

men

t tas

k an

d ac

cide

nt

sequ

ence

mod

ellin

g ta

sk.

RA

TIO

NA

LE: U

se o

f gen

eric

info

rmat

ion

for q

uant

ifica

tion

of th

e re

cove

ry

actio

n fo

r IE

freq

uenc

y es

timat

ion

may

intro

duce

exc

essi

ve

optim

ism

/con

serv

atis

m in

the

resu

lts d

ue to

non

-acc

ount

ing

of p

lant

sp

ecifi

city

:

1) A

bsen

ce/a

vaila

bilit

y of

rele

vant

pro

cedu

res.

2) T

echn

ical

pos

sibi

lity

for r

ecov

ery

actio

n.

3) P

lant

-dep

ende

nt ti

me

mar

gins

for t

he re

cove

ry a

ctio

ns.

EXA

MPL

ES:

1) R

ecov

ery

actio

n fo

r clo

sure

of p

ress

uriz

er sa

fety

val

ves a

fter s

purio

us

open

ing

is d

epen

dent

on

the

actu

al c

ause

of s

purio

us o

peni

ng a

nd d

esig

n of

th

e co

ntro

l circ

uit o

f the

val

ve

2) R

ecov

ery

of th

e of

f-si

te p

ower

dep

ende

nt o

n th

e si

te-s

peci

fic e

xter

nal

grid

cha

ract

eris

tics.

IE-G

11

The

resu

lts o

f the

initi

atin

g ev

ent a

naly

sis a

re c

ompa

red

with

gen

eric

dat

a so

urce

s to

prov

ide

a re

ason

able

ness

che

ck o

f the

qua

ntita

tive

and

qual

itativ

e re

sults

. The

dev

iatio

ns

from

gen

eric

sour

ces a

re re

solv

ed a

nd/o

r exp

lain

ed.

See

CO

MM

ENT

for I

E-G

05.

32

T

able

4.2

-H

Att

ribu

tes f

or IE

Ana

lysi

s: T

ask

IE-H

‘Doc

umen

tatio

n’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

IE-H

D

ocum

enta

tion

and

info

rmat

ion

stor

age

is p

erfo

rmed

in a

man

ner f

acili

tatin

g a

peer

revi

ew,

as w

ell a

s fut

ure

upgr

ades

and

app

licat

ions

of t

he P

SA b

y de

scrib

ing

the

proc

esse

s tha

t w

ere

used

and

pro

vidi

ng d

etai

ls o

f the

met

hods

use

d, a

ssum

ptio

ns m

ade,

and

thei

r bas

es.

IE-H

01

The

follo

win

g is

doc

umen

ted:

I.

IE Id

entif

icat

ion

and

Gro

upin

g:

- IE

def

initi

on.

- Th

e pr

oced

ure

of se

arch

ing

the

spec

ific

initi

atin

g ev

ents

, inc

ludi

ng:

- gen

eric

list

s of I

Es;

- the

ana

lysi

s per

form

ed fo

r sea

rchi

ng p

lant

-uni

que

and

plan

t-spe

cific

initi

ator

s. -

The

appr

oach

for a

sses

sing

the

com

plet

enes

s and

con

sist

ency

of i

nitia

ting

even

ts w

ith

the

plan

t-spe

cific

exp

erie

nce,

indu

stry

exp

erie

nce,

oth

er c

ompa

rabl

e PS

As,

and

gene

ric

initi

atin

g ev

ents

. -

The

ratio

nale

for s

cree

ning

out

initi

ator

s.

- Th

e lis

t of I

Es sc

reen

ed o

ut a

nd sc

reen

ed in

eve

nts.

-

The

basi

s for

gro

upin

g an

d su

bdiv

idin

g in

itiat

ing

even

ts.

- Th

e as

sum

ptio

ns m

ade

to id

entif

y, sc

reen

out

, and

gro

up IE

s. -

The

list o

f IE

grou

ps a

nd p

artic

ular

IEs a

ssig

ned

to.

II. F

requ

enci

es A

sses

smen

t:

- Th

e m

odel

use

d to

eva

luat

e th

e fr

eque

ncy

of e

ach

IE.

- Th

e pr

oces

s for

com

putin

g th

e in

itiat

ing

even

t fre

quen

cies

. -

Sour

ces f

or g

ener

ic e

stim

ates

and

just

ifica

tion

for t

he c

hoic

e of

par

ticul

ar g

ener

ic d

ata

sour

ce(s

). -

The

plan

t-spe

cific

dat

a, in

clud

ing

the

perio

ds, f

or w

hich

pla

nt-s

peci

fic d

ata

wer

e ga

ther

ed fo

r eac

h IE

. -

Just

ifica

tion

for e

xclu

sion

of a

ny d

ata.

-

The

ratio

nale

for a

ny d

istri

butio

ns u

sed

for f

requ

ency

est

imat

ions

. -

Estim

ated

freq

uenc

ies,

incl

udin

g th

e ch

arac

teriz

atio

n of

unc

erta

inty

. -

Pote

ntia

l tim

e de

pend

ent a

spec

ts o

f the

initi

atin

g ev

ent f

requ

enci

es.

- K

ey a

ssum

ptio

ns m

ade

in th

e an

alys

is a

nd th

eir j

ustif

icat

ion

(eng

inee

ring

judg

men

t, sp

ecifi

c an

alys

is, s

tatis

tical

info

rmat

ion,

etc

.).

CO

MM

ENTS

:

1) IE

s fre

quen

cies

mea

n, m

edia

n, 5

% a

nd 9

5% p

erce

ntile

shou

ld b

e as

sess

ed a

nd d

ocum

ente

d.

2) A

ging

of t

he p

lant

shou

ld b

e in

vest

igat

ed fo

r tim

e tre

nds.

33

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

IE-H

02

All

the

unde

rlyin

g da

ta a

nd in

form

atio

n so

urce

s and

ana

lyse

s are

doc

umen

ted

and

stor

ed.

Sp

ecia

l Attr

ibut

e IE

-H02

-S1:

Th

e in

form

atio

n fr

om th

e IE

ana

lysi

s inc

ludi

ng th

e IE

dat

abas

es c

reat

ed

is p

art o

f the

PSA

doc

umen

tatio

n. T

hese

dat

a in

clud

ing

the

data

base

s an

d th

e de

taile

d ba

ckgr

ound

info

rmat

ion

is st

ored

in a

retr

ieva

ble

and

acce

ssib

le e

lect

roni

c fo

rm a

nd fo

rmat

. Due

to th

e am

ount

of

info

rmat

ion

aris

ing

from

the

IE a

naly

sis t

asks

ele

ctro

nic

stor

age

of th

is

info

rmat

ion

is e

ssen

tial f

or m

any

of th

e ap

plic

atio

ns.

CO

MM

ENT:

Ele

ctro

nic

stor

age

of IE

dat

abas

e is

ess

entia

l for

man

y ap

plic

atio

ns.

34

5. PSA ELEMENT ‘AS’: ACCIDENT SEQUENCE ANALYSIS


The objective of the accident sequence (AS) analysis is to ensure that the response of the plant’s systems and operators to an initiating event is reflected in the assessment of CDF in such a way that:

• Significant operator actions, mitigation systems, and phenomena that influence or determine the course of sequences are appropriately included in the accident sequence model and sequence definition.

• Plant-specific dependencies due to initiating events, human interfaces, functional dependencies, environmental, and spatial impact, and common cause failures are reflected in the accident sequence structure.

• The individual function successes, mission times, and time windows for operator actions for each critical safety function modelled in the accident sequences reflects the success criteria evaluated in accordance with the attributes of Section 6 of this publication.

• End states are clearly defined to be either a core damage or successful prevention with the capability to support the interface between Level-1 and Level-2 PSA.

The important aspects of AS analysis are the following:

• Clear definition of success and non-success end states

• Comprehensive list of key safety functions and systems performing the functions

• Realistic accident progression identification

• Clear presentation of AS models

• Completeness of AS models

• Justification for end states for all ASs. 5.2. Accident sequence analysis tasks and their attributes

Table 5.2 lists the main tasks for the PSA element ‘AS Analysis’. Tables 5.2-A through 5.2-E present the description of general and special attributes for these tasks. Table 5.2 Main Tasks for AS Analysis


AS-A Selection of a Method and Provision of Related Tools for Accident Sequences Modelling

AS-B Definition of Success and Non-Success End States and Key Safety Functions

AS-C Accident Sequences Progression Identification and AS Models Development

AS-D Accident Sequence Success Criteria Definition

AS-E Documentation

35

T

able

5.2

-A A

ttri

bute

s for

AS

Ana

lysi

s: T

ask

AS-

A ‘S

elec

tion

of a

Met

hod

and

Prov

isio

n of

Rel

ated

Too

ls fo

r A

ccid

ent S

eque

nces

Mod

ellin

g’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

AS-

A

The

task

incl

udes

the

sele

ctio

n of

a m

etho

d an

d pr

ovis

ion

of re

late

d to

ols f

or a

ccid

ent

sequ

ence

s mod

ellin

g.

AS-

A01

Th

e m

etho

d ch

osen

for A

ccid

ent S

eque

nce

Ana

lysi

s pro

vide

s for

the

poss

ibili

ty to

ex

plic

itly

mod

el th

e ap

prop

riate

com

bina

tions

of s

yste

m re

spon

ses a

nd o

pera

tor a

ctio

ns th

at

affe

ct th

e ke

y sa

fety

func

tions

for e

ach

mod

elle

d in

itiat

ing

even

t /IE

gro

up a

nd p

rovi

des a

fr

amew

ork

to su

ppor

t seq

uenc

e qu

antif

icat

ion.

The

met

hod

supp

orts

gra

phic

al re

pres

enta

tion

of th

e ac

cide

nt se

quen

ce lo

gic

(e.g

. ‘ev

ent

tree

stru

ctur

e’).

RA

TIO

NA

LE: G

raph

ical

repr

esen

tatio

n of

the

AS

logi

c pr

ovid

es fo

r the

po

ssib

ility

to a

naly

se a

nd re

view

AS

mod

els.

This

feat

ure

is o

f hig

h im

porta

nce

for a

num

ber o

f app

licat

ions

.

36

T

able

5.2

-B

Att

ribu

tes f

or A

S A

naly

sis:

Tas

k A

S-B

‘Def

initi

on o

f Suc

cess

and

Non

-Suc

cess

End

Sta

tes a

nd K

ey S

afet

y Fu

nctio

ns’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

AS-

B

For e

ach

initi

atin

g ev

ent g

roup

the

key

safe

ty fu

nctio

ns th

at a

re n

eces

sary

to re

ach

a su

cces

s en

d st

ate

are

iden

tifie

d. S

ucce

ss a

nd n

on-s

ucce

ss e

nd st

ates

are

cle

arly

def

ined

. (Se

e al

so

Tabl

e 6.

2-A

, Tas

k SC

-A).

AS-

B01

Th

e su

cces

s and

non

-suc

cess

stat

es a

re d

efin

ed in

a m

anne

r tha

t pro

vide

s the

pos

sibi

lity

to

just

ify th

e ac

hiev

emen

t (no

n-ac

hiev

emen

t) of

the

succ

ess e

nd st

ate

for e

ach

acci

dent

se

quen

ce w

ith th

e us

e of

ava

ilabl

e to

ols (

e.g.

ther

mal

hyd

raul

ic a

naly

sis,

test

s and

ex

perim

ents

, etc

.). (S

ee a

lso

Tabl

e 6.

2-A

, Gen

eral

Attr

ibut

e SC

-A01

).

All

end

stat

es a

re id

entif

ied

as su

cces

s or n

on-s

ucce

ss; n

o en

d-st

ate

is u

ndet

erm

ined

.

RA

TIO

NA

LE: U

ndef

ined

end

-sta

tes p

reve

nt a

use

ful i

nter

pret

atio

n of

the

resu

lts.

AS-

B02

Fo

r eac

h in

itiat

ing

even

t gro

up th

e ke

y sa

fety

func

tions

are

iden

tifie

d. S

yste

ms a

nd

proc

edur

ally

dire

cted

ope

rato

r act

ions

requ

ired

to p

erfo

rm sa

fety

func

tions

are

iden

tifie

d fo

r ea

ch IE

gro

up w

ith a

ccou

nt fo

r ava

ilabi

lity

of sp

ecifi

c eq

uipm

ent a

nd c

ondi

tions

for

oper

ator

act

ions

(e.g

. inf

orm

atio

n av

aila

ble

for o

pera

tor,

acce

ptab

ility

for m

anua

lly

cont

rolle

d eq

uipm

ent,

time

win

dow

, etc

.).

For e

ach

safe

ty fu

nctio

n, sy

stem

mod

els a

re d

evel

oped

with

acc

ount

for s

ucce

ss c

riter

ia

defin

ed fo

r spe

cific

IE g

roup

and

AS.

(See

als

o Se

ctio

ns 6

and

7).

AS-

B03

A

just

ifica

tion

for t

he a

chie

vem

ent o

f sta

ble

succ

ess e

nd st

ate

cond

ition

s is p

rovi

ded

for

each

AS

with

acc

ount

of a

ll un

certa

intie

s ass

ocia

ted

with

the

appl

icab

le to

ols.

(See

als

o Ta

ble

6.2-

A, G

ener

al A

ttrib

ute

SC-A

02).

RA

TIO

NA

LE: I

gnor

ing

the

unce

rtain

ties a

ssoc

iate

d w

ith th

e av

aila

ble

tool

s use

d to

just

ify a

chie

vem

ent o

f the

succ

ess e

nd st

ate

may

lead

to lo

ss

of si

gnifi

cant

insi

ght o

f the

PSA

.

Sp

ecia

l Attr

ibut

e AS

-B03

-S1:

Ju

stifi

catio

n fo

r the

ach

ieve

men

t of t

he n

on-s

ucce

ss e

nd st

ate

cond

ition

s is

per

form

ed w

ith th

e us

e of

‘bes

t est

imat

e’ m

odel

s and

pa

ram

eter

s of

the

appl

icab

le ju

stifi

catio

n to

ols.

RATI

ON

ALE:

Use

of c

onse

rvat

ive

para

met

ers f

or ju

stifi

catio

n of

non

-su

cces

s end

stat

es m

ay le

ad to

exc

essi

vely

con

serv

ativ

e co

nsid

erat

ion

of

cert

ain

ASs,

bias

the

resu

lts a

nd in

sigh

ts a

nd m

ake

cert

ain

appl

icat

ions

no

n-cr

edib

le.

37

T

able

5.2

-C

Att

ribu

tes f

or A

S A

naly

sis:

Tas

k A

S-C

‘Acc

iden

t Seq

uenc

es P

rogr

essi

on Id

entif

icat

ion

and

AS

Mod

els D

evel

opm

ent’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

AS-

C

For e

ach

IE g

roup

the

acci

dent

pro

gres

sion

for a

ll se

quen

ces i

s ide

ntifi

ed a

nd ju

stifi

ed. F

or

each

IE g

roup

the

acci

dent

sequ

ence

mod

els a

re d

evel

oped

. AS

mod

els e

xplic

itly

addr

ess

real

istic

pla

nt b

ehav

iour

in re

spon

se to

IE in

term

s of n

orm

al p

lant

syst

ems o

pera

tion,

op

erat

or a

ctio

ns, a

nd m

itiga

tion

syst

ems t

hat s

uppo

rt th

e ke

y sa

fety

func

tions

nec

essa

ry to

ac

hiev

e a

stab

le sa

fe st

ate.

AS-

C01

Th

e en

d st

ates

of t

he a

ccid

ent s

eque

nces

are

ach

ieve

d w

hen

eith

er a

‘non

-suc

cess

ful’

or

‘sta

ble

succ

essf

ul’ p

lant

con

ditio

ns h

ave

been

reac

hed.

C

OM

MEN

T: ‘S

tabl

e su

cces

sful

’ pla

nt c

ondi

tions

are

the

cond

ition

s w

hich

may

be

mai

ntai

ned

durin

g an

d af

ter t

he d

efin

ed m

issi

on ti

me

with

the

set o

f equ

ipm

ent p

ostu

late

d to

be

oper

able

for t

he sp

ecifi

c ac

cide

nt se

quen

ce. I

n ca

se st

able

succ

essf

ul p

lant

con

ditio

ns fo

r se

quen

ces a

re n

ot a

chie

ved

with

in 2

4 ho

urs,

a lo

nger

mis

sion

tim

e is

to

be c

onsi

dere

d an

d/or

add

ition

al p

lant

equ

ipm

ent a

nd h

uman

inte

ract

ions

ar

e m

odel

led.

EXA

MPL

E: T

he p

ossi

bilit

y to

rem

ove

heat

via

the

seco

ndar

y si

de u

sing

lim

ited

amou

nt o

f wat

er in

the

dem

iner

alis

ed w

ater

tank

s for

ap

prox

imat

ely

24 h

ours

shou

ld n

ot b

e co

nsid

ered

as a

succ

essf

ul st

able

en

d st

ate.

AS-

C02

R

ealis

tic a

nd ‘a

pplic

able

’ (i.e

., fr

om ‘s

imila

r’ p

lant

s) th

erm

al h

ydra

ulic

ana

lyse

s are

use

d to

de

term

ine

the

acci

dent

pro

gres

sion

par

amet

ers (

e.g.

tim

ing,

tem

pera

ture

, pre

ssur

e). A

ll no

rmal

ope

ratio

n an

d st

and-

by sy

stem

s, th

e op

erab

ility

of w

hich

may

impa

ct th

e ac

cide

nt

prog

ress

ion

are

acco

unte

d fo

r.

Sp

ecia

l Attr

ibut

e AS

-C02

-S1:

Plan

t-spe

cific

real

istic

ther

mal

hyd

raul

ic a

naly

ses a

re u

sed

to

dete

rmin

e th

e ac

cide

nt p

rogr

essi

on fo

r ASs

. RA

TIO

NAL

E : U

se o

f the

rmal

hyd

raul

ic a

naly

sis f

rom

sim

ilar u

nits

may

pr

oduc

e re

sults

, whi

ch d

o no

t acc

ount

for s

peci

fic p

lant

feat

ures

in

fluen

cing

acc

iden

t pro

gres

sion

for s

peci

fic se

quen

ces.

The

ASs

cons

truc

ted

with

out t

akin

g in

to a

ccou

nt p

lant

-spe

cific

feat

ures

may

not

be

app

ropr

iate

for s

ome

PSA

appl

icat

ions

.

Con

serv

ativ

e as

sum

ptio

ns a

re m

ade

whe

n pa

rticu

lar c

ours

e of

acc

iden

t pro

gres

sion

for

spec

ific

ASs

are

not

just

ified

by

supp

ortin

g an

alys

es (e

.g. t

herm

al h

ydra

ulic

, fra

ctur

al

mec

hani

cs, r

eact

ivity

ana

lysi

s, et

c.).

RA

TIO

NA

LE: T

his a

ttrib

ute

is st

ated

in o

rder

to a

void

mis

sing

of

pote

ntia

l ins

ight

due

to la

ck o

f kno

wle

dge

on a

ctua

l pla

nt b

ehav

iour

. If

safe

ty si

gnifi

cant

insi

ghts

can

not b

e ac

hiev

ed w

ith th

e us

e of

co

nser

vativ

e as

sum

ptio

ns, m

ore

effo

rts sh

ould

be

take

n to

rem

ove

cons

erva

tism

with

app

ropr

iate

just

ifica

tion.

AS-

C03

Spec

ial A

ttrib

ute

AS-C

03-S

1:

Plan

t-spe

cific

real

istic

ana

lyse

s are

per

form

ed fo

r spe

cific

ASs

in o

rder

to

obt

ain

a re

alis

tic d

escr

iptio

n an

d m

odel

of a

ccid

ent s

eque

nces

. RA

TIO

NAL

E : U

se o

f con

serv

ativ

e as

sum

ptio

ns in

stea

d of

real

istic

an

alys

is m

ay b

ias t

he b

enef

its o

f cer

tain

app

licat

ions

aim

ed a

t im

prov

ing/

chec

king

the

influ

ence

of s

peci

fic p

lant

cha

nges

.

38

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

The

proc

edur

es a

re re

view

ed w

ith e

ngag

emen

t of p

lant

ope

ratio

ns a

nd tr

aini

ng p

erso

nnel

to

conf

irm th

at th

e in

terp

reta

tion

of th

e pr

oced

ures

and

the

expe

cted

resp

onse

s are

con

sist

ent

with

the

exis

ting

ther

mal

hyd

raul

ic a

naly

ses a

nd p

lant

ope

ratio

nal p

ract

ices

.

The

acci

dent

sequ

ence

mod

els a

re c

onsi

sten

t with

the

plan

t-spe

cific

em

erge

ncy

proc

edur

es,

train

ing

sim

ulat

or e

xerc

ises

, and

exi

stin

g th

erm

al h

ydra

ulic

ana

lyse

s. In

cas

e of

alte

rnat

ives

, th

e m

ost r

estri

ctiv

e ac

cide

nt p

rogr

essi

on is

mod

elle

d.

(See

als

o Ta

ble

8.2-

E, G

ener

al A

ttrib

utes

HR

-E01

, HR

-E02

, and

Tab

le 8

.2-G

, Gen

eral

A

ttrib

ute

HR

-G04

).

A

S-C

04

Spec

ial A

ttrib

ute

AS-C

04-S

1:

Whe

n ex

istin

g pr

oced

ures

allo

w o

pera

tor t

o fo

llow

diff

eren

t miti

gatio

n st

rate

gies

, the

acc

iden

t seq

uenc

e m

odel

s acc

ount

for a

ll po

ssib

le

stra

tegi

es.

The

likel

ihoo

d of

eac

h st

rate

gy is

est

imat

ed b

ased

on

the

resu

lts o

f int

ervi

ew w

ith p

lant

ope

rato

rs, a

ctua

l pla

nt e

xper

ienc

e, a

nd

trai

ning

pra

ctic

e.

RATI

ON

ALE:

Inc

ompl

ete

mod

ellin

g of

acc

iden

t pro

gres

sion

dea

ling

with

non

-cle

ar re

quir

emen

ts o

f pla

nt e

mer

genc

y pr

oced

ures

may

bia

s th

e be

nefit

s of c

erta

in a

pplic

atio

ns a

imed

at i

mpr

ovin

g th

e ac

cide

nt

proc

edur

es.

For a

ll ac

cide

nt se

quen

ces c

onst

ruct

ed b

ased

on

the

resu

lts o

f rea

listic

ther

mal

hyd

raul

ic

anal

yses

, whi

ch re

quire

hum

an in

tera

ctio

ns n

ot c

onsi

dere

d in

pla

nt e

mer

genc

y pr

oced

ures

for

the

parti

cula

r sce

nario

, the

failu

re o

f tho

se h

uman

inte

ract

ions

is a

ssum

ed o

r add

ition

al

inve

stig

atio

ns a

re p

erfo

rmed

to a

ssur

e th

e po

ssib

ility

to p

erfo

rm re

quire

d ac

tions

(e.g

. pla

nt -

spec

ific

ther

mal

hyd

raul

ic a

naly

sis f

or sp

ecifi

c se

quen

ces,

inte

rvie

ws w

ith p

lant

ope

rato

rs,

anal

ysis

of p

lant

exp

erie

nce,

sim

ulat

or e

xerc

ises

, etc

.)

(See

als

o Ta

ble

8.2-

E, G

ener

al A

ttrib

utes

HR

-E01

, HR

-E02

, and

Tab

le 8

.2-G

, Gen

eral

A

ttrib

ute

HR

-G04

).

RA

TIO

NA

LE:

Opt

imis

tic c

redi

ting

the

hum

an in

tera

ctio

ns n

ot

desc

ribed

in th

e em

erge

ncy

proc

edur

es m

ay m

ask

the

prob

lem

s in

the

emer

genc

y pr

oced

ures

.

CO

MM

ENT:

Em

erge

ncy

proc

edur

es m

ay b

e le

ss d

etai

led

than

acc

iden

t se

quen

ce m

odel

s in

the

PSA

.

AS-

C05

Spec

ial A

ttrib

ute

AS-C

05-S

1:

For a

ll ac

cide

nt se

quen

ces c

onst

ruct

ed b

ased

on

the

resu

lts o

f rea

listic

th

erm

al h

ydra

ulic

ana

lyse

s, w

hich

requ

ire

hum

an in

tera

ctio

ns n

ot

cons

ider

ed in

pla

nt e

mer

genc

y pr

oced

ures

for t

he p

artic

ular

scen

ario

, ad

ditio

nal i

nves

tigat

ions

are

per

form

ed to

ass

ure

the

poss

ibili

ty to

pe

rfor

m re

quir

ed a

ctio

ns (e

.g. p

lant

-spe

cific

ther

mal

hyd

raul

ic a

naly

sis

for s

peci

fic se

quen

ces,

inte

rvie

ws w

ith p

lant

ope

rato

rs, a

naly

sis o

f pla

nt

expe

rien

ce, s

imul

ator

exe

rcis

es, e

tc.)

RATI

ON

ALE:

Rea

listic

mod

ellin

g of

acc

iden

t pro

gres

sion

s dea

ling

with

no

n-cl

ear r

equi

rem

ents

of p

lant

em

erge

ncy

proc

edur

es h

elps

to a

sses

s re

al b

enef

its fr

om im

prov

emen

ts in

em

erge

ncy

oper

atin

g pr

oced

ures

.

AS-

C06

If e

mer

genc

y pr

oced

ures

per

mit

perf

orm

ing

an a

ctio

n, b

ut th

e re

alis

tic t/

h an

alys

es

dem

onst

rate

that

the

actio

n im

pose

s an

adve

rse

effe

ct, t

he A

Ss n

ever

thel

ess i

nclu

de th

ose

actio

ns. (

See

also

Tab

le 8

.2-E

, Gen

eral

Attr

ibut

es H

R-E

01, H

R-E

02, a

nd T

able

8.2

-G,

Gen

eral

Attr

ibut

e H

R-G

04).

RA

TIO

NA

LE: U

njus

tifie

d op

timis

m sh

ould

be

avoi

ded

in A

Ss

mod

ellin

g.

39

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

Sp

ecia

l Attr

ibut

e AS

-C06

-S1:

If em

erge

ncy

proc

edur

es p

erm

it to

per

form

an

actio

n bu

t the

real

istic

t/h

ana

lyse

s dem

onst

rate

that

the

actio

n im

pose

s an

adve

rse

effe

ct,

addi

tiona

l inv

estig

atio

ns a

re p

erfo

rmed

to a

sses

s the

real

ope

rato

r re

spon

se (

e.g.

int

ervi

ews w

ith p

lant

ope

rato

rs, a

naly

sis o

f pla

nt

expe

rien

ce, s

imul

ator

exe

rcis

es, e

tc.).

RATI

ON

ALE:

Rea

listic

mod

ellin

g of

acc

iden

t pro

gres

sion

s dea

ling

with

no

n-cl

ear r

equi

rem

ents

of p

lant

em

erge

ncy

proc

edur

es h

elps

to a

sses

s re

al b

enef

its fr

om im

prov

emen

ts in

em

erge

ncy

oper

atin

g pr

oced

ures

.

AS-

C07

Th

e ac

cide

nt se

quen

ces p

ossi

ble

for e

ach

initi

atin

g ev

ent g

roup

are

dev

elop

ed. F

or e

ach

initi

atin

g ev

ent g

roup

, the

acc

iden

t seq

uenc

es m

odel

is d

evel

oped

to a

suff

icie

nt le

vel o

f de

tail

that

the

diff

eren

ces i

n re

quire

men

ts o

n sy

stem

s and

ope

rato

r res

pons

es a

re c

aptu

red.

D

iver

se sy

stem

s pro

vidi

ng si

mila

r fun

ctio

n ar

e m

odel

led

sepa

rate

ly if

the

choi

ce o

f one

sy

stem

ove

r ano

ther

sign

ifica

ntly

cha

nges

the

requ

irem

ents

for o

pera

tor i

nter

vent

ion

or th

e ne

ed fo

r oth

er sy

stem

s.

CO

MM

ENT:

Acc

iden

t seq

uenc

es c

an b

e m

odel

led

at v

ario

us le

vels

of

deta

il ra

ngin

g fr

om s

mal

l fun

ctio

nal l

evel

eve

nt tr

ees,

thro

ugh

syst

em

leve

l eve

nt tr

ees (

com

mon

ly re

ferr

ed to

as t

he sm

all e

vent

tree

app

roac

h w

ith fa

ult t

ree

linki

ng) t

o ve

ry la

rge

even

t tre

es th

at m

odel

supp

ort

syst

em a

nd fr

ont l

ine

syst

em s

tatu

s at t

he tr

ain

leve

l (th

e so

-cal

led

‘larg

e ev

ent t

ree’

or ‘

even

t tre

e lin

king

’ app

roac

h). (

See

also

Tab

le 1

1.2-

A,

Gen

eral

Attr

ibut

e M

Q-A

01).

AS-

C08

Fo

r eac

h ke

y sa

fety

func

tion

its d

epen

denc

e on

the

succ

ess o

r fai

lure

of p

rece

ding

func

tions

an

d th

e im

pact

on

acci

dent

pro

gres

sion

are

add

ress

ed.

EX

AM

PLE:

The

succ

ess o

f low

pre

ssur

e sy

stem

inje

ctio

n is

dep

ende

nt

on th

e su

cces

s of R

PV d

epre

ssur

isat

ion.

AS-

C09

W

hen

deve

lopi

ng a

ccid

ent s

eque

nces

, the

phe

nom

enol

ogic

al c

ondi

tions

cre

ated

by

the

acci

dent

pro

gres

sion

are

iden

tifie

d so

that

the

effe

ct o

n po

tent

ial m

itiga

ting

syst

ems i

s pr

oper

ly a

ccou

nted

for.

(See

als

o Ta

ble

10.2

-C, G

ener

al A

ttrib

ute

DF-

C01

).

EXA

MPL

E: P

heno

men

olog

ical

con

ditio

ns in

clud

e ge

nera

tion

of h

arsh

en

viro

nmen

tal e

ffec

ts, i

nclu

ding

tem

pera

ture

, pre

ssur

e, d

ebris

, wat

er

leve

ls, a

nd h

umid

ity.

The

effe

cts o

f the

se c

ondi

tions

cou

ld d

irect

ly

caus

e eq

uipm

ent f

ailu

re, o

r mig

ht re

quire

pro

cedu

ral o

pera

tor a

ctio

ns to

pr

even

t equ

ipm

ent d

amag

e (e

.g. t

o te

mpo

rary

dis

able

equ

ipm

ent).

AS-

C10

If

pla

nt c

onfig

urat

ions

and

mai

nten

ance

pra

ctic

es c

reat

e de

pend

enci

es a

mon

g va

rious

sys

tem

al

ignm

ents

, the

se c

onfig

urat

ions

and

alig

nmen

ts a

re d

efin

ed a

nd in

clud

ed in

the

mod

el in

a

man

ner t

hat r

efle

cts t

hese

dep

ende

ncie

s.

AS-

C11

Ev

ents

for w

hich

tim

e-ph

ased

dep

ende

ncie

s mig

ht e

xist

are

def

ined

and

are

incl

uded

in A

Ss

mod

els a

ppro

pria

tely

. EX

AM

PLES

of t

ime

phas

ed e

vent

s inc

lude

: AC

pow

er re

cove

ry, D

C

batte

ry ti

me

depe

nden

t dis

char

ge, e

nviro

nmen

tal c

ondi

tions

for

oper

atin

g eq

uipm

ent a

nd th

e co

ntro

l roo

m (e

.g. r

oom

coo

ling)

, etc

.

AS-

C12

If

ATW

S se

quen

ces a

re g

roup

ed a

nd m

odel

led

as a

sing

le e

vent

tree

, the

mos

t res

trict

ive

case

is

con

side

red.

C

OM

MEN

T: M

any

PSA

s gro

up A

TWS

sequ

ence

s tog

ethe

r and

de

velo

p a

mod

el fo

r the

mos

t res

trict

ive

case

(e.g

. LO

OP,

etc

.)

Sp

ecia

l Attr

ibut

e AS

-C

12-S

1 AT

WS

sequ

ence

s are

mod

elle

d as

par

t of a

ccid

ent s

eque

nce

mod

el fo

r ea

ch IE

gro

up. S

peci

fic a

naly

ses j

ustif

y th

e po

ssib

ility

to p

reve

nt c

ore

dam

age

for e

ach

ATW

S se

quen

ce w

ith su

cces

sful

end

stat

e.

CO

MM

ENT:

For

mod

ellin

g of

ATW

S se

quen

ces t

he sp

ecifi

c co

nditi

ons

whi

ch a

re in

pla

ce fo

r the

rela

ted

initi

atin

g ev

ent g

roup

are

acc

ount

ed

for,

as w

ell a

s the

con

ditio

ns a

nd e

vent

s con

nect

ed to

a p

artic

ular

fa

ilure

of t

he re

acto

r shu

tdow

n fu

nctio

n.

40

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

AS-

C13

W

hen

trans

fers

bet

wee

n ev

ent t

rees

are

use

d to

redu

ce th

e si

ze a

nd c

ompl

exity

of i

ndiv

idua

l ev

ent t

rees

the

met

hod

for i

mpl

emen

ting

an e

vent

tree

tran

sfer

that

pre

serv

es th

e de

pend

enci

es th

at a

re p

art o

f the

tran

sfer

red

sequ

ence

is u

sed.

The

se in

clud

e fu

nctio

nal,

syst

em, i

nitia

ting

even

t, op

erat

or, a

nd sp

atia

l or e

nviro

nmen

tal d

epen

denc

ies.

CO

MM

ENT:

If so

ftwar

e be

ing

used

for P

SA is

not

cap

able

of e

vent

tre

es li

nkin

g, c

are

mus

t be

take

n th

at a

ll bo

unda

ry c

ondi

tions

of p

aren

t ev

ent t

ree

top

even

t are

tran

sfer

red

to su

bseq

uent

eve

nt tr

ees.

The

cons

eque

nces

of s

ucce

ssfu

l ope

ratio

n of

miti

gatin

g sy

stem

s on

acci

dent

pro

gres

sion

are

de

term

ined

by

cons

ider

ing

the

actu

al p

lant

resp

onse

.

EXA

MPL

ES:

1) O

n re

ceip

t of a

LO

CA

sign

al, t

hree

con

tain

men

t spr

ay p

umps

star

t ev

en th

ough

onl

y on

e is

requ

ired

by th

e su

cces

s crit

eria

. If

ther

e is

no

dire

ctiv

e to

the

oper

ator

s to

turn

off

pum

ps, o

r dec

reas

e flo

w, t

he

incr

ease

d flo

w w

ill d

ecre

ase

the

time

of d

eple

tion

of th

e ta

nk sh

ared

by

com

mon

spay

and

inje

ctio

n sy

stem

s. E

ven

if th

ere

is p

roce

dura

l di

rect

ion

to d

ecre

ase

flow

, the

pos

sibi

lity

that

the

oper

ator

s wou

ld fa

il to

do

so sh

ould

be

take

n in

to a

ccou

nt.

2) O

peni

ng o

f all

stea

m sa

fety

val

ves w

hile

onl

y on

e is

nee

ded

to

prev

ent e

xces

sive

pre

ssur

e in

crea

se in

the

stea

m li

nes m

ay le

ad to

fa

ilure

to re

-clo

se o

f sev

eral

val

ves.

AS-

C14

Spec

ial A

ttrib

ute

AS-C

14-S

1:

Real

istic

pla

nt-s

peci

fic a

naly

ses a

re m

ade

for s

peci

fic A

Ss in

ord

er to

ve

rify

whe

ther

the

cond

ition

s for

ope

rato

r act

ions

and

ope

ratio

n of

the

spec

ific

equi

pmen

t are

ach

ieve

d. (S

ee a

lso

Tabl

e 8.

2-E,

Spe

cial

At

trib

ute

HR-

E01-

S1).

RATI

ON

ALE:

Use

of c

onse

rvat

ive

assu

mpt

ions

inst

ead

of re

alis

tic

anal

ysis

may

bia

s the

ben

efits

of m

any

appl

icat

ions

aim

ed a

t im

prov

ing/

chec

king

the

influ

ence

of s

peci

fic p

lant

cha

nges

(e.g

. ha

rdw

are

or p

roce

dure

s).

AS-

C15

A

ccid

ent p

rogr

essi

on is

dis

cuss

ed a

nd ‘a

gree

d’ w

ith p

lant

ope

rato

rs.

RA

TIO

NA

LE: E

xper

ienc

ed p

lant

ope

rato

rs c

ould

iden

tify

inco

nsis

tenc

ies b

etw

een

PSA

ASs

mod

els a

nd a

ctua

l pla

nt b

ehav

iour

in

term

of a

ccid

ent p

rogr

essi

on, s

yste

m re

quire

men

ts, p

roce

dura

lly

dire

cted

ope

rato

r act

ions

, etc

.

Sp

ecia

l Attr

ibut

e AS

-C15

-S1:

An

exp

ande

d gr

aphi

cal r

epre

sent

atio

n of

acc

iden

ts p

rogr

essi

on (e

.g.

‘eve

nt se

quen

ce d

iagr

am’)

for I

E gr

oups

is u

sed

to v

erify

the

AS m

odel

s w

ith p

lant

ope

rato

rs.

RATI

ON

ALE:

Gra

phic

al re

pres

enta

tion

of a

ccid

ent p

rogr

essi

on h

elps

no

n-PS

A sp

ecia

lists

to u

nder

stan

d AS

mod

els.

41

T

able

5.2

-D

Att

ribu

tes f

or A

S A

naly

sis:

Tas

k A

S-D

‘Suc

cess

Cri

teri

a fo

r A

ccid

ent S

eque

nces

’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

AS-

D

The

task

incl

udes

the

defin

ition

of a

ccid

ent s

eque

nce

succ

ess c

riter

ia.

AS-

D01

Fo

r eac

h in

itiat

ing

even

t gro

up a

nd fo

r eac

h ac

cide

nt se

quen

ce, t

he su

cces

s crit

eria

for

safe

ty re

late

d fu

nctio

ns, o

pera

tor a

ctio

ns, s

yste

ms,

and

equi

pmen

t are

def

ined

. (Se

e de

taile

d at

tribu

tes i

n Se

ctio

n 6)

.

42

T

able

5.2

-E

Att

ribu

tes f

or A

S A

naly

sis:

Tas

k A

S-E

‘Doc

umen

tatio

n’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

AS-

E

Doc

umen

tatio

n an

d in

form

atio

n st

orag

e is

per

form

ed in

a m

anne

r tha

t fac

ilita

tes p

eer

revi

ew, a

s wel

l as f

utur

e up

grad

es a

nd a

pplic

atio

ns o

f the

PSA

by

desc

ribin

g th

e pr

oces

ses

that

wer

e us

ed a

nd p

rovi

ding

det

ails

of t

he m

etho

ds u

sed,

ass

umpt

ions

mad

e an

d th

eir

base

s.

AS-

E01

The

treat

men

t of e

ach

initi

ator

and

acc

iden

t seq

uenc

e m

odel

is d

ocum

ente

d to

supp

ort

revi

ews a

nd a

pplic

atio

ns.

The

follo

win

g im

porta

nt a

spec

ts a

re d

ocum

ente

d:

- G

raph

ical

repr

esen

tatio

n of

eac

h ac

cide

nt se

quen

ce fo

r eac

h IE

gro

up;

- A

des

crip

tion

of th

e ac

cide

nt p

rogr

essi

on fo

r eac

h se

quen

ce o

r gro

up o

f sim

ilar

sequ

ence

s;

- Th

e su

cces

s crit

eria

est

ablis

hed

for e

ach

initi

atin

g ev

ent c

ateg

ory

incl

udin

g th

e ba

ses

for t

he c

riter

ia;

- A

ny a

ssum

ptio

ns th

at w

ere

mad

e in

dev

elop

ing

the

acci

dent

sequ

ence

s, as

wel

l as t

he

base

s for

the

assu

mpt

ions

; -

Exis

ting

anal

yses

per

form

ed to

def

ine

succ

ess c

riter

ia a

nd e

xpec

ted

sequ

ence

ph

enom

ena

incl

udin

g ne

cess

ary

timin

g co

nsid

erat

ions

; -

Suff

icie

nt sy

stem

ope

ratio

n in

form

atio

n to

supp

ort t

he m

odel

led

depe

nden

cies

; -

Cal

cula

tions

or o

ther

bas

es u

sed

to ju

stify

equ

ipm

ent o

pera

bilit

y be

yond

its ‘

norm

al’

desi

gn p

aram

eter

s and

for w

hich

cre

dit h

as b

een

take

n;

- Ju

stifi

catio

n fo

r the

non

-loss

of d

epen

denc

es if

mod

ellin

g si

mpl

ifica

tions

wer

e im

plie

d.

AS-

E02

The

inte

rfac

es b

etw

een

Acc

iden

t Seq

uenc

e A

naly

sis a

nd th

e fo

llow

ing

PSA

task

s are

de

fined

and

doc

umen

ted:

-

The

defin

ition

of i

nitia

ting

even

t cat

egor

y in

the

Initi

atin

g Ev

ent A

naly

sis T

ask;

-

The

defin

ition

of c

ore

dam

age

and

asso

ciat

ed su

cces

s crit

eria

in S

ucce

ss C

riter

ia

Def

initi

on T

ask;

-

Key

def

initi

ons o

f ope

rato

r act

ions

and

sequ

ence

-spe

cific

tim

ing

and

depe

nden

cies

re

flect

ed in

the

acci

dent

sequ

ence

mod

els i

n th

e H

RA

task

for t

hese

act

ions

; -

The

basi

s for

the

sequ

ence

and

cut

set q

uant

ifica

tion

in th

e Le

vel-1

Qua

ntifi

catio

n an

d R

esul

ts In

terp

reta

tion

Task

; -

A fr

amew

ork

for a

n in

tegr

ated

trea

tmen

t of d

epen

denc

ies i

n th

e in

itiat

ing

even

ts

anal

ysis

, sys

tem

s ana

lysi

s, da

ta a

naly

sis,

hum

an re

liabi

lity

anal

ysis

, Lev

el-1

qu

antif

icat

ion.

43

6. PSA ELEMENT ‘SC’: SUCCESS CRITERIA FORMULATION AND SUPPORTING ANALYSIS


The main objective of the success criteria formulation task is to determine for given initiating events what represents a successful or unsuccessful plant response and to translate this information into detailed plant system and operator action success criteria. Thermal hydraulic analyses simulating the course of accident sequence progression and other assessment means are used for this purpose. These analyses and assessments are called in this section supporting analyses for the success criteria formulation.

As a first task core or fuel damage or other unsuccessful accident sequence end states

are defined in order to provide the basis for the derivation of detailed success criteria for safety related functions or human interactions. The description of the formulation of success criteria and of related attributes in this section is limited to success criteria required for a Level-1 PSA.

Success criteria regarding the plant response to initiating events are used to specify

whether safety related functions meet the requirements to prevent damage to the core or mitigate significant releases of radioactivity. These safety-related functions in terms of a PSA may be functions of operating systems, front line safety systems, I&C, support systems, structures, components, and operator actions. For operator actions success criteria are characterized by statements that certain actions are successfully carried out within a defined time window.

Success criteria are used to construct the logic PSA model, including for example the

determination of event tree branch point probabilities and probabilities for other events in the logic model. They also determine the required number of trains of a safety related system. Top-level safety related function requirements are translated into the requirements for systems performing that function, a process, which is continued down to support systems. Event tree branch point probabilities also may reflect operator actions with their specific success criteria, e.g. a time window. Other operator actions may be modelled at the system level. There is therefore a close connection between HRA (Human Reliability Analysis), systems analysis, and success criteria formulation.

6.2. Success criteria formulation and supporting analysis tasks and their attributes

Table 6.2 lists the main tasks for the PSA element ‘Success Criteria Formulation and

Supporting Analysis’. Tables 6.2-A through 6.2-C present the description of general and special attributes for these tasks.

44

Table 6.2 Main Tasks for Success Criteria Formulation and Supporting Analysis


SC-A Definition of Overall and Detailed Success Criteria

SC-B Thermal Hydraulic Analyses and other Assessment Means Supporting the Derivation of Detailed Success Criteria

SC-C Documentation

45

T

able

6.2

-A

Att

ribu

tes f

or S

ucce

ss C

rite

ria

Form

ulat

ion

and

Supp

ortin

g A

naly

sis:

Tas

k SC

-A ‘D

efin

ition

of O

vera

ll an

d D

etai

led

Succ

ess

Cri

teri

a’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

SC-A

D

efin

ition

of t

he o

vera

ll su

cces

s crit

eria

for t

he P

SA a

nd d

efin

ition

of t

he su

cces

s crit

eria

fo

r sys

tem

s, st

ruct

ures

, com

pone

nts,

and

hum

an in

tera

ctio

ns is

per

form

ed in

a m

anne

r w

hich

is c

onsi

sten

t with

pla

nt fe

atur

es, p

roce

dure

s and

ope

ratin

g pr

actic

es.

SC-A

01

Def

initi

on o

f cor

e da

mag

e or

oth

er u

nsuc

cess

ful a

ccid

ent s

eque

nce

end

stat

es

Cor

e or

fuel

dam

age

is d

efin

ed in

term

s of p

hysi

cal p

roce

sses

and

phe

nom

ena

and

failu

re

mec

hani

sms,

whi

ch m

ay c

ause

a su

bsta

ntia

l rel

ease

of r

adio

activ

e m

ater

ial f

rom

the

reac

tor f

uel.

The

follo

win

g in

form

atio

n so

urce

s are

use

d to

der

ive

the

defin

ition

of c

ore

or fu

el

dam

age:

- Th

e av

aila

ble

desi

gn b

asis

info

rmat

ion

for a

pla

nt a

nd re

late

d in

form

atio

n, e

.g. d

esig

n ba

sis a

ccid

ent a

naly

ses i

s use

d.

- A

vaila

ble

info

rmat

ion

on c

ore

or fu

el d

amag

e m

echa

nism

s and

phe

nom

ena

for s

imila

r pl

ants

and

from

rela

ted

expe

rimen

tal i

nves

tigat

ions

is u

sed.

CO

MM

ENT:

For

ves

sel t

ype

LWR

s the

re a

re u

sual

ly o

nly

two

type

s of

end-

stat

es d

efin

ed in

term

s of t

he L

evel

-1 P

SA:

(1) S

ucce

ssfu

l end

-sta

tes w

ithou

t sig

nific

ant c

ore

dam

age,

and

(2) U

nsuc

cess

ful e

nd-s

tate

s with

cor

e da

mag

e.

For o

ther

reac

tor t

ypes

, for

exa

mpl

e fo

r cha

nnel

type

reac

tors

, diff

eren

t le

vels

of c

ore

or fu

el d

amag

e ar

e us

ed to

refle

ct sc

enar

ios w

here

dam

age

is li

mite

d to

onl

y on

e ch

anne

l, a

grou

p of

cha

nnel

s, to

a p

ortio

n of

the

core

or e

xten

ds to

the

entir

e co

re.

SC-A

02

The

phys

ical

pla

nt p

aram

eter

s (e.

g. h

ighe

st n

ode

tem

pera

ture

, cor

e co

llaps

ed li

quid

leve

l) an

d as

soci

ated

acc

epta

nce

crite

ria o

r lim

it va

lues

(e.g

. tem

pera

ture

lim

it, p

erce

ntag

e of

cl

addi

ng th

ickn

ess o

xidi

zed)

to b

e us

ed in

det

erm

inin

g co

re o

r fue

l dam

age

are

defin

ed.

The

para

met

ers a

re se

lect

ed in

such

a w

ay th

at th

e de

term

inat

ion

of c

ore

or fu

el d

amag

e is

as

real

istic

as p

ract

ical

and

con

sist

ent w

ith c

urre

nt b

est p

ract

ices

and

kno

wle

dge.

For

the

appl

icat

ion

of p

aram

eter

acc

epta

nce

crite

ria w

ith th

e re

sults

of t

herm

al h

ydra

ulic

ca

lcul

atio

n a

suff

icie

nt m

argi

n is

spec

ified

and

use

d to

take

car

e of

lim

itatio

ns o

f the

co

mpu

ter c

odes

, suc

h as

lim

itatio

ns in

the

soph

istic

atio

n of

mod

els,

and

unce

rtain

ties i

n th

e re

sults

.

EXA

MPL

ES o

f par

amet

ers a

nd a

ssoc

iate

d ac

cept

ance

crit

eria

that

are

us

ed in

PSA

s inc

lude

:

- BW

R: C

olla

psed

liqu

id le

vel l

ess t

han

1/3

core

hei

ght o

r cod

e-pr

edic

ted

peak

cor

e te

mpe

ratu

re >

250

0°F

(137

0°C

)

- PW

R: C

olla

psed

liqu

id le

vel b

elow

top

of a

ctiv

e fu

el fo

r a p

rolo

nged

pe

riod,

or c

ode-

pred

icte

d co

re p

eak

node

tem

pera

ture

> 2

200°

F (1

200°

C)

usin

g a

code

with

det

aile

d co

re m

odel

ling,

or c

ode-

pred

icte

d co

re p

eak

node

tem

pera

ture

> 1

800°

F (1

000°

C) u

sing

a c

ode

with

sim

plifi

ed (e

.g.

sing

le-n

ode

core

mod

el, l

umpe

d pa

ram

eter

) cor

e m

odel

ling,

or c

ode-

pred

icte

d co

re e

xit t

empe

ratu

re >

120

0°F

(650

°C) f

or 3

0 m

in u

sing

a c

ode

with

sim

plifi

ed c

ore

mod

ellin

g.

SC-A

03

Succ

ess c

riter

ia fo

r eac

h of

the

safe

ty re

late

d fu

nctio

n ar

e sp

ecifi

ed in

the

acci

dent

se

quen

ces f

or e

ach

initi

atin

g ev

ent g

roup

. C

OM

MEN

T: T

he fo

rmul

atio

n of

succ

ess c

riter

ia a

t the

func

tiona

l lev

el,

syst

em le

vel,

syst

em tr

ain

leve

l and

for s

truct

ures

and

equ

ipm

ent i

s ac

com

pani

ed b

y a

desc

riptio

n of

acc

iden

t seq

uenc

es, i

nclu

ding

refe

renc

es

to th

e pl

ants

pro

tect

ive

I&C

and

EO

Ps w

hich

prim

arily

det

erm

ine

the

plan

t sys

tem

s res

pons

e.

46

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

SC-A

04

Syst

ems c

apab

le o

f mee

ting

the

spec

ified

succ

ess c

riter

ia o

f saf

ety

rela

ted

func

tions

are

id

entif

ied

toge

ther

with

ass

ocia

ted

oper

ator

act

ions

if a

pplic

able

. The

succ

ess c

riter

ia fo

r th

e sa

fety

rela

ted

func

tions

are

tran

slat

ed a

ccor

ding

ly to

the

syst

ems a

nd a

ssoc

iate

d op

erat

or a

ctio

ns.

From

the

leve

l of s

yste

ms,

succ

ess c

riter

ia a

re c

ontin

ued

to sy

stem

trai

ns if

nec

essa

ry a

nd

furth

er d

own

to th

e as

soci

ated

supp

ort s

yste

ms.

For m

ultip

le u

nits

pla

nts t

he sy

stem

s tha

t are

shar

ed b

etw

een

units

are

iden

tifie

d, a

nd th

e m

anne

r in

whi

ch th

e sh

arin

g is

per

form

ed sh

ould

the

units

exp

erie

nce

a co

mm

on in

itiat

ing

even

t (e.

g. L

OO

P). T

his i

nclu

des o

pera

tor a

ctio

ns a

s req

uire

d an

d sp

ecifi

ed b

y pl

ant

proc

edur

es o

r ope

ratin

g pr

actic

es.

CO

MM

ENT:

As s

peci

fied

unde

r oth

er P

SA ta

sks i

n th

is p

ublic

atio

n de

pend

enci

es o

f fro

nt li

ne sy

stem

s on

supp

ort s

yste

ms a

nd I&

C a

re

pref

erab

ly m

odel

led

expl

icitl

y fo

r exa

mpl

e by

mea

ns o

f tra

nsfe

rs fr

om

supp

ort s

yste

m tr

ains

to th

e eq

uipm

ent o

f the

fron

t lin

e sy

stem

whi

ch

depe

nd o

n a

parti

cula

r sup

port

syst

ems.

Dep

endi

ng o

n th

e de

sign

this

re

quire

s for

mul

atio

n of

succ

ess c

riter

ia fo

r sup

port

syst

ems a

s wel

l. A

n ex

ampl

e fo

r thi

s is a

DC

ele

ctric

al p

ower

supp

ly fr

om tw

o D

C su

pply

tra

ins v

ia se

para

tion

diod

es. A

part

from

thes

e st

raig

htfo

rwar

d tra

in

depe

nden

cies

on

supp

orts

ther

e ar

e us

ually

dep

ende

ncie

s req

uirin

g ad

ditio

nal a

naly

ses,

for e

xam

ple

room

coo

ling

requ

irem

ents

rega

rdin

g th

e op

erat

ion

of sa

fety

rela

ted

equi

pmen

t dur

ing

the

mis

sion

tim

e.

SC-A

05

A m

issi

on ti

me

for t

he a

ccid

ent s

eque

nces

is d

eter

min

ed. T

he m

issi

on ti

me

is th

e tim

e du

ring

whi

ch sa

fety

rela

ted

func

tions

are

requ

ired

to w

ork

to a

chie

ve a

stab

le e

nd-s

tate

af

ter a

n in

itiat

ing

even

t. It

is u

sual

pra

ctic

e to

ass

ume

as a

firs

t app

roac

h a

gene

ral m

issi

on

time

of 2

4 ho

urs.

For s

eque

nces

in w

hich

stab

le p

lant

con

ditio

ns w

ould

not

be

achi

eved

by

24 h

ours

usi

ng th

e m

odel

led

plan

t equ

ipm

ent a

nd h

uman

inte

ract

ions

, a lo

nger

mis

sion

tim

e is

use

d if

need

ed to

ach

ieve

stab

le p

lant

con

ditio

ns. (

See

also

AS-

02).

Add

ition

al e

valu

atio

n or

mod

ellin

g is

car

ried

out f

or se

quen

ces i

n w

hich

a sa

fe, s

tabl

e st

ate

has n

ot b

een

achi

eved

by

the

end

of th

e m

issi

on ti

me

defin

ed fo

r the

PSA

by

usin

g te

chni

ques

like

:

a) a

ssig

ning

an

appr

opria

te p

lant

dam

age

stat

e fo

r tha

t seq

uenc

e if

this

is u

sefu

l and

doe

s no

t sig

nific

antly

hin

der a

pplic

atio

ns;

b) e

xten

ding

the

mis

sion

tim

e, a

nd e

xten

ding

the

affe

cted

ana

lyse

s to

the

poin

t at w

hich

co

nditi

ons a

nd p

aram

eter

s can

be

show

n to

reac

h ac

cept

able

val

ues;

c) m

odel

ling

addi

tiona

l sys

tem

reco

very

or o

pera

tor i

nter

actio

ns fo

r the

sequ

ence

, in

acco

rdan

ce w

ith re

quire

men

ts st

ated

in th

e sy

stem

s ana

lysi

s and

HR

A se

ctio

ns o

f thi

s gu

ide,

to d

emon

stra

te th

at a

succ

essf

ul o

utco

me

is a

chie

ved

toge

ther

with

an

asse

ssm

ent o

f the

pro

babi

lity

for s

ucce

ss a

nd fa

ilure

of t

he a

dditi

onal

eve

nts.

47

T

able

6.2

-B

Att

ribu

tes f

or S

ucce

ss C

rite

ria

Form

ulat

ion

and

Supp

ortin

g A

naly

sis:

Tas

k SC

-B ‘T

herm

al H

ydra

ulic

Ana

lyse

s and

Oth

er

Ass

essm

ent M

eans

Sup

port

ing

the

Form

ulat

ion

of S

ucce

ss C

rite

ria’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

SC-B

D

etai

led

succ

ess c

riter

ia a

nd e

vent

tim

ing

suff

icie

nt fo

r the

qua

ntifi

catio

n of

CD

F, a

nd th

e de

term

inat

ion

of im

pact

of s

ucce

ss c

riter

ia o

n sy

stem

s, st

ruct

ures

, com

pone

nts a

nd h

uman

in

tera

ctio

ns a

re d

evel

oped

by

appr

opria

te th

erm

al h

ydra

ulic

ana

lyse

s and

oth

er a

sses

smen

t m

eans

.

SC-B

01

App

licab

le a

nd p

rove

n co

mpu

ter c

odes

are

use

d fo

r the

mod

ellin

g of

the

cour

se o

f acc

iden

t se

quen

ces a

nd fo

r the

der

ivat

ion

of a

ssoc

iate

d su

cces

s crit

eria

. Pre

fera

bly

and

if av

aila

ble

best

est

imat

e co

des a

nd m

odel

s are

use

d fo

r thi

s pur

pose

and

the

plan

t and

sequ

ence

mod

el

refle

cts t

he sp

ecifi

c de

sign

and

ope

ratio

nal f

eatu

res o

f the

pla

nt.

Bes

t-est

imat

e m

odel

s or a

naly

ses c

an b

e su

pple

men

ted

with

pla

nt sp

ecifi

c or

gen

eric

FSA

R

or o

ther

con

serv

ativ

e an

alys

es a

pplic

able

to th

e pl

ant a

ccom

pani

ed w

ith a

just

ifica

tion

and

an a

sses

smen

t of a

ssoc

iate

d un

certa

intie

s.

Sp

ecia

l Attr

ibut

e SC

-B01

-S1:

Appl

icab

le a

nd p

rove

n co

mpu

ter c

odes

are

use

d fo

r the

mod

ellin

g th

e co

urse

of a

ll re

leva

nt a

ccid

ent s

eque

nces

and

for t

he d

eriv

atio

n of

as

soci

ated

succ

ess c

rite

ria.

Bes

t est

imat

e co

des a

nd m

odel

s are

use

d fo

r th

is p

urpo

se a

nd th

e pl

ant a

nd se

quen

ce m

odel

refle

cts t

he sp

ecifi

c de

sign

and

ope

ratio

nal f

eatu

res o

f the

pla

nt.

RATI

ON

ALE:

Use

of g

ener

ic a

sses

smen

ts m

ay p

rovi

de in

suffi

cien

t pla

nt

spec

ifici

ty fo

r par

ts o

f the

mod

el a

ffect

ed b

y ap

plic

atio

ns. C

onse

rvat

ive

asse

ssm

ents

may

cau

se m

aski

ng e

ffect

s hin

deri

ng c

erta

in a

pplic

atio

ns.

SC-B

02

Expe

rt ju

dgm

ent i

s onl

y us

ed to

ass

ess t

he c

ondi

tions

or r

espo

nse

of sy

stem

s, st

ruct

ures

and

eq

uipm

ent i

n si

tuat

ions

whe

n th

ere

is a

lack

of a

vaila

ble

info

rmat

ion,

kno

wle

dge

or

anal

ytic

al m

etho

ds u

pon

whi

ch a

pre

dict

ion

can

be b

ased

if it

can

be

dem

onst

rate

d th

at th

e va

riabi

lity

and

unce

rtain

ty p

oten

tially

inhe

rent

in th

e as

sess

men

t doe

s not

sign

ifica

ntly

im

pact

on

the

PSA

mod

els a

nd re

sults

.

SC-B

03

Whe

n de

finin

g su

cces

s crit

eria

, the

rmal

hyd

raul

ic, s

truct

ural

, or o

ther

ana

lyse

s and

ev

alua

tions

are

use

d w

hich

are

app

ropr

iate

to th

e ev

ent a

nd th

e ev

ent s

eque

nce

bein

g an

alys

ed. T

he le

vel o

f det

ail i

n th

ese

anal

yses

and

eva

luat

ions

is c

onsi

sten

t with

the

initi

atin

g ev

ent g

roup

ing

and

acci

dent

sequ

ence

ana

lysi

s tas

ks.

48

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

SC-B

04

Ana

lysi

s mod

els a

nd c

ompu

ter c

odes

are

use

d th

at h

ave

suff

icie

nt c

apab

ility

to m

odel

the

cond

ition

s and

phe

nom

ena

of in

tere

st in

the

dete

rmin

atio

n of

succ

ess c

riter

ia a

nd th

at

prov

ide

resu

lts re

pres

enta

tive

for t

he p

lant

. Com

pute

r cod

es a

nd m

odel

s are

onl

y us

ed

with

in k

now

n lim

its o

f app

licab

ility

.

SC-B

05

The

plan

t mod

el a

nd p

aram

eter

s use

d fo

r the

rmal

hyd

raul

ic a

naly

ses a

re e

stab

lishe

d in

a

way

that

pro

vide

s suf

ficie

nt re

solu

tion

and

refle

cts t

he a

ctua

l des

ign

and

oper

atio

nal

feat

ures

of t

he p

lant

.

Para

met

er v

alue

s (in

clud

ing

setp

oint

s, lim

it po

ints

, trig

ger v

alue

s, en

try a

nd e

xit v

alue

s for

pr

oced

ures

, and

sets

of p

aram

eter

val

ues w

hich

are

use

d fo

r con

trol f

unct

ions

) det

erm

ine

whe

n op

erat

or a

ctio

ns a

re c

arrie

d ou

t and

det

erm

ine

the

func

tion

of sa

fety

rela

ted

syst

ems.

In th

is re

spec

t the

pla

nt m

odel

and

the

mod

el o

f rel

ated

con

trol s

yste

m fu

nctio

ns a

re

supp

orte

d by

a d

etai

led

desc

riptio

n in

clud

ing

refe

renc

es to

the

plan

ts p

rote

ctiv

e I&

C a

nd

EOPs

.

Whe

n sp

ecify

ing

para

met

er v

alue

s, un

certa

intie

s, va

riabi

lity,

and

del

ays f

or m

easu

ring

and

actu

atin

g de

vice

s and

for a

ctua

ted

equi

pmen

t are

take

n in

to a

ccou

nt.

SC-B

06

The

plau

sibi

lity,

reas

onab

lene

ss, a

nd a

ccep

tabi

lity

of th

erm

al h

ydra

ulic

, stru

ctur

al o

r oth

er

supp

ortin

g en

gine

erin

g ba

ses u

sed

to su

ppor

t suc

cess

crit

eria

is c

heck

ed w

ith a

ppro

pria

te

met

hods

, suc

h as

:

a)

Com

paris

on o

f res

ults

with

resu

lts o

f sim

ilar a

naly

ses p

erfo

rmed

for s

imila

r pla

nts,

acco

untin

g fo

r diff

eren

ces i

n un

ique

pla

nt fe

atur

es.

b)

Com

paris

on w

ith re

sults

of s

imila

r ana

lyse

s with

oth

er c

odes

.

c)

Che

ck b

y ot

her m

eans

, e.g

. sim

plifi

ed e

ngin

eerin

g ca

lcul

atio

ns.

49

T

able

6.2

-C

Att

ribu

tes f

or S

ucce

ss C

rite

ria

Form

ulat

ion

and

Supp

ortin

g A

naly

sis:

Tas

k SC

-C ‘D

ocum

enta

tion’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

SC-C

Th

e do

cum

enta

tion

of su

cces

s crit

eria

form

ulat

ion

and

of th

e su

ppor

ting

anal

ysis

is c

arrie

d ou

t in

a m

anne

r, w

hich

is fu

lly tr

acea

ble

and

allo

ws c

hang

ing

and

repr

oduc

tion

of th

e as

sess

men

t if r

equi

red

by a

n ap

plic

atio

n.

SC-C

01

Each

of t

he su

cces

s crit

eria

and

the

supp

ortin

g as

sess

men

ts, e

ngin

eerin

g ba

ses,

refe

renc

es

and

assu

mpt

ions

are

doc

umen

ted

in d

etai

l:

- C

onse

rvat

ive

assu

mpt

ions

are

des

crib

ed a

nd d

ocum

ente

d in

clud

ing

the

ratio

nale

for

usin

g co

nser

vatis

m a

nd a

n as

sess

men

t of i

mpa

cts.

- A

det

aile

d an

d tra

ceab

le d

escr

iptio

n of

con

dens

atio

n, g

roup

ing,

bin

ning

, ag

glom

erat

ion,

scre

enin

g an

d si

mpl

ifica

tion

step

s is g

iven

incl

udin

g ju

stifi

catio

ns a

nd

an a

sses

smen

t of e

ffect

s and

whi

ch is

con

sist

ent w

ith th

e in

itiat

ing

even

t and

acc

iden

t se

quen

ce a

naly

sis t

asks

.

- Th

e ba

sis f

or th

e su

cces

s crit

eria

dev

elop

men

t pro

cess

is d

ocum

ente

d in

a w

ay, w

hich

is

con

sist

ent w

ith th

e in

itiat

ing

even

t and

acc

iden

t seq

uenc

e an

alys

is ta

sks.

SC-C

02

The

uses

, rat

iona

le a

nd b

ackg

roun

d in

form

atio

n fo

r exp

ert j

udgm

ent i

s doc

umen

ted

in a

w

ay w

hich

is tr

acea

ble

and

repr

oduc

ible

.

Unc

erta

intie

s and

var

iabi

litie

s in

expe

rt ju

dgm

ent a

re st

ated

. The

impa

cts o

r effe

cts o

f the

se

unce

rtain

ties a

nd v

aria

bilit

ies a

re a

sses

sed

as p

art o

f the

Res

ults

Ana

lysi

s and

Inte

rpre

tatio

n Ta

sk.

(See

als

o Ta

ble

12.2

-B, G

ener

al A

ttrib

ute

RI-

B02

).

RA

TIO

NA

LE:

App

licat

ions

may

requ

ire re

doin

g ex

pert

judg

men

ts,

parti

ally

or a

s a w

hole

. Mea

ning

ful r

esul

ts a

nd c

ompa

rison

s can

onl

y be

ac

hiev

ed in

such

cas

e if

the

expe

rt ju

dgm

ent p

roce

ss in

clud

ing

back

grou

nd in

form

atio

n an

d ju

stifi

catio

ns a

re su

ffic

ient

ly d

ocum

ente

d.

SC-C

03

The

ratio

nale

use

d in

the

appl

icat

ion

of su

cces

s crit

eria

for s

ituat

ions

for w

hich

ther

e is

m

ore

than

one

tech

nica

l app

roac

h, n

one

of w

hich

is u

nive

rsal

ly a

ccep

ted

as c

orre

ct, i

s do

cum

ente

d an

d th

e ef

fect

s of u

sing

a p

artic

ular

app

roac

h is

just

ified

and

impa

cts

disc

usse

d.

SC-C

04

The

follo

win

g is

doc

umen

ted

cons

iste

nt w

ith a

nd n

ot e

xten

sive

ly d

oubl

ing

the

info

rmat

ion

pres

ente

d in

the

task

s for

initi

atin

g ev

ents

, acc

iden

t seq

uenc

e as

sess

men

t, hu

man

inte

ract

ion

anal

ysis

and

syst

ems a

naly

sis:

a)

The

defin

ition

of c

ore

dam

age

used

in th

e PS

A in

clud

ing

the

basi

s for

any

sele

cted

pa

ram

eter

s and

par

amet

er v

alue

s.

b)

Cal

cula

tions

(gen

eric

and

pla

nt-s

peci

fic) o

r oth

er re

fere

nces

use

d to

est

ablis

h su

cces

s cr

iteria

, and

iden

tific

atio

n of

cas

es fo

r whi

ch th

ey a

re u

sed.

c)

Iden

tific

atio

n of

com

pute

r cod

es o

r oth

er m

etho

ds u

sed

to e

stab

lish

plan

t-spe

cific

su

cces

s crit

eria

.

50

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

d)

A d

escr

iptio

n of

the

limita

tions

(e.g

. pot

entia

l con

serv

atis

ms o

r lim

itatio

ns th

at c

ould

ch

alle

nge

the

appl

icab

ility

of c

ompu

ter c

ode

mod

els i

n ce

rtain

cas

es) o

f the

ca

lcul

atio

ns o

r cod

es.

e)

Iden

tific

atio

n of

impo

rtant

ass

umpt

ions

use

d in

est

ablis

hing

succ

ess c

riter

ia.

f)

A d

etai

led

and

trace

able

des

crip

tion

of c

onde

nsat

ion,

gro

upin

g, b

inni

ng,

aggl

omer

atio

n, sc

reen

ing-

out a

nd si

mpl

ifica

tion

step

s whe

re u

sed

incl

udin

g ju

stifi

catio

ns a

nd im

pact

ass

essm

ent a

nd c

onsi

sten

t with

the

initi

atin

g ev

ent a

nd

acci

dent

sequ

ence

ana

lysi

s tas

ks.

g)

A su

mm

ary

of su

cces

s crit

eria

for t

he sa

fety

rela

ted

func

tions

, sys

tem

s and

hum

an

inte

ract

ions

for e

ach

acci

dent

initi

atin

g ev

ent g

roup

.

h)

The

basi

s for

det

erm

inin

g th

e tim

e w

indo

ws a

nd o

ther

con

ditio

ns fo

r hum

an

inte

ract

ions

.

i) Th

e de

scrip

tion

of p

roce

sses

use

d to

def

ine

succ

ess c

riter

ia fo

r gro

uped

initi

atin

g ev

ents

or a

ccid

ent s

eque

nces

.

51

7. PSA ELEMENT ‘SY’: SYSTEMS ANALYSIS

7.1 Main objectives The objectives of the systems analysis element are to identify and quantify the causes

of failure for each plant system represented in the initiating event analysis and accident sequence analysis in such a way that:

• For each safety function in accident sequence models, system models are developed with account for success criteria.

• System-level success criteria, mission times, time windows for operator actions, different initial system alignments and assumptions provide the basis for the system logic models as reflected in the model. A reasonably complete set of system failure and unavailability modes for each system is represented.

• Human errors and operator actions that could influence the system unavailability or the system’s contribution to accident sequences are identified for development as part of the HRA element.

• Intersystem dependencies and intra-system dependencies including functional, human, phenomenological, and common-cause failures that could influence system unavailability or the system’s contribution to accident-sequence frequencies are identified and accounted for.

7.2. Systems analysis tasks and their attributes Table 7.2 lists the main tasks for the PSA element ‘Systems Analysis’. Tables 7.2-A

through 7.2-D present the description of general and special attributes for these tasks.

Table 7.2 Main Tasks for Systems Analysis


SY-A System Characterisation and System Boundary Definition

SY-B Failure Cause Identification and Modelling

SY-C Identification and Modelling of Dependencies

SY-D Documentation

52

T

able

7.2

-A

Att

ribu

tes f

or S

yste

ms A

naly

sis:

Tas

k SY

-A ‘S

yste

m C

hara

cter

isat

ion

and

Syst

em B

ound

ary

Def

initi

on’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

SY-A

Sy

stem

cha

ract

eris

tics i

nclu

ding

bou

ndar

ies a

re d

efin

ed fo

r all

syst

ems,

incl

udin

g su

ppor

t sy

stem

s, ne

eded

for p

erfo

rmin

g th

e fu

nctio

ns id

entif

ied

in th

e ac

cide

nt se

quen

ce a

naly

sis.

SY-A

01

Plan

t inf

orm

atio

n so

urce

s are

revi

ewed

in o

rder

to:

- D

efin

e sy

stem

func

tion

durin

g no

rmal

and

acc

iden

t con

ditio

ns

- Es

tabl

ish

syst

em b

ound

arie

s

- Id

entif

y in

terf

aces

with

oth

er sy

stem

s

- Id

entif

y in

stru

men

tatio

n an

d co

ntro

l req

uire

men

ts in

clud

ing

oper

ator

inte

rfac

e

- Id

entif

y te

stin

g an

d m

aint

enan

ce re

quire

men

ts a

nd p

ract

ices

- Id

entif

y op

erat

ing

limita

tions

such

as t

hose

impo

sed

by te

chni

cal s

peci

ficat

ions

- Id

entif

y pr

oced

ures

for t

he o

pera

tion

of th

e sy

stem

dur

ing

norm

al a

nd a

ccid

ent

cond

ition

s

- Id

entif

y sy

stem

con

figur

atio

n du

ring

norm

al a

nd a

ccid

ent c

ondi

tions

- Id

entif

y sy

stem

test

and

surv

eilla

nce

proc

edur

es

- A

scer

tain

syst

em o

pera

ting

hist

ory

- A

scer

tain

sys

tem

mod

ifica

tion

hist

ory.

EXA

MPL

ES o

f inf

orm

atio

n so

urce

s inc

lude

:

Syst

em P

&ID

s, on

e-lin

e di

agra

ms,

inst

rum

enta

tion

and

cont

rol

draw

ings

, spa

tial l

ayou

t dra

win

gs, s

yste

m o

pera

ting

proc

edur

es,

abno

rmal

ope

ratin

g pr

oced

ures

, em

erge

ncy

proc

edur

es, s

ucce

ss

crite

ria c

alcu

latio

ns, t

he fi

nal o

r upd

ated

SA

R, t

echn

ical

sp

ecifi

catio

ns, t

rain

ing

info

rmat

ion,

syst

em d

escr

iptio

ns a

nd re

late

d de

sign

doc

umen

ts, a

ctua

l sys

tem

ope

ratin

g ex

perie

nce

and

inte

rvie

ws

with

syst

em e

ngin

eers

and

ope

rato

rs.

SY-A

02

Com

pone

nts r

equi

red

for s

yste

m o

pera

tion

and

the

supp

ort s

yste

ms i

nter

face

s req

uire

d fo

r ac

tuat

ion

and

oper

atio

n of

the

syst

em c

ompo

nent

s are

iden

tifie

d.

CO

MM

ENT:

The

bou

ndar

ies o

f sys

tem

s def

ined

in a

PSA

may

be

diffe

rent

from

the

boun

dary

use

d in

the

plan

t bei

ng a

naly

sed.

53

T

able

7.2

-B

Att

ribu

tes f

or S

yste

ms A

naly

sis:

Tas

k SY

-B: ‘

Failu

re C

ause

Iden

tific

atio

n an

d M

odel

ling’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

SY-B

Sy

stem

mod

els a

re d

evel

oped

for a

ll sy

stem

s inc

lude

d in

Tas

k A

.

SY-B

01

The

syst

em m

odel

s inc

lude

with

in th

e bo

unda

ry th

e co

mpo

nent

s req

uire

d fo

r sys

tem

op

erat

ion

(as i

dent

ified

abo

ve),

incl

udin

g in

terf

aces

as i

dent

ified

in S

Y-A

02 (T

able

7.2

-A).

SY-B

02

Bot

h no

rmal

and

alte

rnat

e sy

stem

alig

nmen

ts a

re m

odel

led.

C

OM

MEN

T: D

epen

ding

on

the

mod

ellin

g te

chni

que,

a si

ngle

sy

stem

mod

el m

ay b

e co

nstru

cted

that

add

ress

es a

ll al

ignm

ents

, or

sepa

rate

mod

els m

ay b

e de

velo

ped

for e

ach

diffe

rent

alig

nmen

t.

SY-B

03

Syst

em m

odel

s are

dev

elop

ed fo

r all

succ

ess c

riter

ia re

quire

d in

the

acci

dent

sequ

ence

m

odel

s.

COM

MEN

TS:

1) S

ucce

ss c

riter

ia fo

r all

syste

ms a

re d

evel

oped

acc

ordi

ng to

Sec

tion

6.

2) D

epen

ding

on

the

mod

ellin

g te

chni

que,

a si

ngle

syste

m m

odel

may

be

con

struc

ted

that

add

ress

es a

ll su

cces

s crit

eria

, or s

epar

ate

mod

els

may

be

deve

lope

d fo

r eac

h su

cces

s crit

erio

n.

EXA

MPL

E: (a

) diff

eren

t suc

cess

crit

eria

are

requ

ired

for s

ome

syste

ms t

o m

itiga

te d

iffer

ent a

ccid

ent s

cena

rios:

the

num

ber o

f pum

ps

requ

ired

to o

pera

te in

som

e sy

stem

s is d

epen

dent

upo

n th

e ac

cide

nt

initi

atin

g ev

ent;

(b) s

ucce

ss c

riter

ia fo

r som

e sy

stem

s are

dep

ende

nt o

n th

e su

cces

s of a

noth

er c

ompo

nent

in th

e sy

stem

; (c)

succ

ess c

riter

ia

for s

ome

syste

ms a

re ti

me-

depe

nden

t.

SY-B

04

The

boun

darie

s of t

he c

ompo

nent

s req

uire

d fo

r sys

tem

ope

ratio

n m

atch

the

defin

ition

s use

d to

es

tabl

ish th

e co

mpo

nent

failu

re d

ata.

Thi

s attr

ibut

e co

mpl

ies a

lso w

ith G

ener

al A

ttrib

ute

SY-

C02

(Tab

le 7

.2-C

).

EXA

MPL

E: A

loca

l con

trol c

ircui

t for

a p

ump

does

not

nee

d to

be

incl

uded

exp

licitl

y in

the

syste

m m

odel

if th

e co

ntro

l circ

uit i

s not

sh

ared

with

ano

ther

com

pone

nt a

nd th

e pu

mp

failu

re d

ata

used

in

quan

tifyi

ng th

e sy

stem

mod

el in

clud

e co

ntro

l circ

uit f

ailu

res.

SY-B

05

A sy

stem

atic

met

hod

is u

sed

for i

dent

ifica

tion

of c

ompo

nent

una

vaila

bilit

y an

d fa

ilure

m

odes

. EX

AM

PLE

is th

e us

e of

FM

EA.

54

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

Syst

ems m

odel

s are

dev

elop

ed to

incl

ude

all c

ompo

nent

failu

re m

odes

and

una

vaila

bilit

ies

that

lead

to fa

ilure

to a

chie

ve sy

stem

func

tion

as d

efin

ed b

y th

e sy

stem

succ

ess c

riter

ia e

xcep

t as

exc

lude

d by

SY

-B13

.

EXA

MPL

E of

failu

re m

odes

(not

a c

ompr

ehen

sive

list

): -

activ

e co

mpo

nent

fails

to st

art;

- ac

tive

com

pone

nt fa

ils to

con

tinue

to ru

n;

- fa

ilure

of a

clo

sed

com

pone

nt to

ope

n;

- fa

ilure

of a

clo

sed

com

pone

nt to

rem

ain

clos

ed;

- fa

ilure

of a

n op

en c

ompo

nent

to c

lose

; -

failu

re o

f an

open

com

pone

nt to

rem

ain

open

; -

activ

e co

mpo

nent

spur

ious

ope

ratio

n;

- pl

uggi

ng o

f an

activ

e or

pas

sive

com

pone

nt;

- le

akag

e of

an

activ

e or

pas

sive

com

pone

nt;

- ru

ptur

e of

an

activ

e or

pas

sive

com

pone

nt;

- in

tern

al le

akag

e of

a c

ompo

nent

; -

inte

rnal

rupt

ure

of a

com

pone

nt;

- fa

ilure

to p

rovi

de si

gnal

(e.g

. ins

trum

enta

tion)

; -

spur

ious

sign

al/o

pera

tion;

-

pre-

initi

ator

hum

an fa

ilure

eve

nts3 .

SY-B

06

Spec

ial A

ttrib

ute

SY-B

06-S

1 Le

akag

es/r

uptu

res o

f pas

sive

com

pone

nts a

re in

clud

ed in

the

mod

el.

RATI

ON

ALE :

Mod

ellin

g of

pas

sive

com

pone

nts f

ailu

res (

e.g.

pip

e se

gmen

ts) i

s use

ful f

or P

SA a

pplic

atio

ns d

ealin

g w

ith o

ptim

izat

ion

of

ISI.

SY-B

07

The

mod

els i

nclu

de c

onsi

dera

tion

to fl

ow d

iver

sion

as a

resu

lt of

com

pone

nt fa

ilure

s whe

n th

e flo

w d

iver

sion

is su

ffici

ent t

o fa

il a

syst

em o

r a tr

ain

func

tion

as d

efin

ed b

y th

e sy

stem

su

cces

s crit

eria

. Ex

clus

ion

of fl

ow d

iver

sion

failu

re m

odes

from

the

mod

el is

just

ified

by

anal

ysis

.

EXA

MPL

ES a

re:

- Fl

ow is

div

erte

d th

roug

h re

circ

ulat

ion

lines

. -

Flow

is d

iver

ted

by ro

und

pum

ping

due

to p

ump

failu

re a

nd

failu

re o

f che

ck v

alve

to re

-clo

se.

- Fl

ow is

div

erte

d th

roug

h sp

urio

usly

ope

n ov

erpr

essu

re

prot

ectio

n sa

fety

relie

f val

ve in

the

syst

em.

3 ‘Pre

-initi

ator

hum

an fa

ilure

eve

nt’ r

epre

sent

s the

failu

re o

f pla

nt st

aff t

o pe

rform

cor

rect

ly th

e re

quire

d ac

tiviti

es th

at c

ause

s the

una

vaila

bilit

y of

the

com

pone

nt, s

yste

m, o

r fun

ctio

n. T

hese

ac

tiviti

es a

re u

sual

ly d

ealin

g w

ith te

st a

nd m

aint

enan

ce.

55

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

SY-B

08

Una

vaila

bilit

y fo

r com

pone

nts d

ue to

test

ing

and

mai

nten

ance

(bot

h pr

even

tive

and

corr

ectiv

e) is

incl

uded

in th

e sy

stem

mod

els c

onsi

sten

t with

the

actu

al p

ract

ices

and

his

tory

of

the

plan

t for

rem

ovin

g eq

uipm

ent f

rom

serv

ice:

-

unav

aila

bilit

y ca

used

by

test

ing

whe

n a

com

pone

nt o

r sys

tem

trai

n is

reco

nfig

ured

from

its

requ

ired

acci

dent

miti

gatin

g po

sitio

n su

ch th

at th

e co

mpo

nent

can

not f

unct

ion

as

requ

ired

if an

initi

atin

g ev

ent o

ccur

; -

mai

nten

ance

eve

nts a

t the

trai

n le

vel w

hen

proc

edur

es re

quire

isol

atin

g th

e en

tire

train

fo

r mai

nten

ance

; -

mai

nten

ance

eve

nts a

t a su

b-tra

in le

vel (

i.e.,

betw

een

tag

out b

ound

arie

s, su

ch a

s a

func

tiona

l equ

ipm

ent g

roup

) whe

n di

rect

ed b

y pr

oced

ures

. R

estri

ctio

ns o

n co

inci

dent

una

vaila

bilit

y of

com

pone

nts/

train

s due

to m

aint

enan

ce b

ased

on

plan

t adm

inis

trativ

e pr

actic

es su

ch a

s Tec

hnic

al S

peci

ficat

ions

are

add

ress

ed.

EXA

MPL

ES o

f out

-of-

serv

ice

unav

aila

bilit

y to

be

mod

elle

d:

- Tr

ain

outa

ges d

urin

g a

wor

k w

indo

w fo

r pre

vent

ive/

corr

ectiv

e m

aint

enan

ce

- A

func

tiona

l equ

ipm

ent g

roup

rem

oved

from

serv

ice

for

prev

entiv

e/co

rrec

tive

mai

nten

ance

-

A re

lief v

alve

take

n ou

t of s

ervi

ce

CO

MM

ENT:

The

coi

ncid

ent m

aint

enan

ce o

n tw

o re

dund

ant t

rain

s, if

perm

itted

, may

be

an im

porta

nt ri

sk c

ontri

buto

r, an

d th

is

poss

ibili

ty n

eeds

to b

e co

nsid

ered

.

Sp

ecia

l Attr

ibut

e SY

-B08

-S1:

Th

e sy

stem

mod

els i

nclu

de th

e ab

ility

to tu

rn te

st a

nd m

aint

enan

ce

cont

ribu

tions

on

and

off.

RATI

ON

ALE :

E.g

. a R

isk

Mon

itor r

efle

cts a

ctua

l con

figur

atio

n an

d m

aint

enan

ce a

ctiv

ities

and

mea

n va

lue

mai

nten

ance

una

vaila

bilit

y ar

e no

t use

d.

Sp

ecia

l Attr

ibut

e SY

-B08

-S2:

Sy

stem

mod

els r

efle

ct th

e re

al m

aint

enan

ce si

tuat

ion

with

mai

nten

ance

un

avai

labi

lity

attr

ibut

ed to

eac

h tr

ain.

Attr

ibut

ing

all m

aint

enan

ce

unav

aila

bilit

y to

one

par

ticul

ar tr

ain

is a

void

ed.

RATI

ON

ALE:

Cer

tain

app

licat

ions

will

requ

ire

trai

n-sp

ecifi

c m

odel

ling.

Sp

ecia

l Attr

ibut

e SY

-B08

-S3:

Sy

mm

etri

c m

odel

s are

dev

elop

ed to

avo

id o

vere

stim

atio

n of

impo

rtan

ce o

f so

me

part

icul

ar re

dund

ant c

ompo

nent

s or t

rain

s and

und

eres

timat

ion

of

othe

rs. A

ttrib

utin

g th

e IE

loca

lizat

ion

or c

halle

nges

for e

quip

men

t ac

tuat

ion

to p

artic

ular

com

pone

nts f

rom

the

set o

f red

unda

nt c

ompo

nent

s is

avo

ided

.

RATI

ON

ALE:

The

app

licat

ions

dea

ling

with

rank

ing

com

pone

nts i

n ac

cord

ance

with

thei

r im

port

ance

mea

sure

s req

uire

equ

al

cons

ider

atio

n of

the

poss

ibili

ty o

f an

IE to

occ

ur in

any

of t

he

redu

ndan

t tra

ins o

r loo

ps a

nd o

f the

redu

ndan

t com

pone

nts t

o op

erat

e w

hen

dem

ande

d.

EXAM

PLE:

N

on-s

ymm

etri

c m

odel

: Ste

am li

ne ru

ptur

e is

mod

eled

as a

rupt

ure

on a

sing

le se

lect

ed S

G (e

.g. S

G1)

with

the

tota

l IE

freq

uenc

y at

trib

uted

to th

e st

eam

line

ass

ocia

ted

with

SG

1; in

this

cas

e th

e ef

fect

of u

nava

ilabi

lity

of th

e SG

1 is

olat

ion

valv

e w

ill b

e ov

eres

timat

ed a

nd th

e ef

fect

of u

nava

ilabi

litie

s of e

quiv

alen

t is

olat

ion

valv

es o

n ot

her S

Gs u

nder

estim

ated

. Sy

mm

etri

c m

odel

: Ste

am li

ne ru

ptur

e is

mod

eled

as a

rupt

ure

on a

ny

SG w

ith th

e fr

eque

ncy

part

ition

ed e

qual

ly b

etw

een

all S

Gs.

In th

is

case

, im

port

ance

mea

sure

s of r

edun

dant

com

pone

nts a

re n

ot b

iase

d.

56

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

SY-B

09

In th

e sy

stem

s ana

lysi

s, al

l hum

an e

rror

s tha

t cau

se th

e sy

stem

or c

ompo

nent

to b

e un

avai

labl

e w

hen

dem

ande

d ar

e in

clud

ed (s

ee T

asks

HR

-A th

roug

h H

R-C

in S

ectio

n 8)

. EX

AM

PLE:

An

oper

ator

fails

to re

stor

e a

com

pone

nt to

its c

orre

ct

stat

e af

ter m

aint

enan

ce a

nd e

rron

eous

cal

ibra

tion.

Tas

k an

alys

is is

an

exa

mpl

e of

a m

etho

d th

at c

an b

e us

ed to

iden

tify

pote

ntia

l err

ors.

SY-B

10

In th

e sy

stem

mod

el, o

pera

tor e

rror

s tha

t can

occ

ur d

urin

g th

e op

erat

ion

of th

e sy

stem

or

com

pone

nts a

re in

clud

ed.

CO

MM

ENT:

Ope

rato

r err

ors n

eed

not b

e in

clud

ed in

the

syst

em

logi

c if

they

are

alre

ady

incl

uded

in th

e ac

cide

nt se

quen

ce m

odel

s w

here

the

rele

vant

func

tion

of th

e sy

stem

is m

odel

led

(see

Tas

ks

HR

-E a

nd H

R-F

in S

ectio

n 8)

.

SY-B

11

Com

pone

nt fa

ilure

s tha

t wou

ld b

e be

nefic

ial t

o sy

stem

ope

ratio

n ar

e no

t inc

lude

d in

the

mod

el.

EXA

MPL

E of

a b

enef

icia

l fai

lure

: A fa

ilure

of a

n in

stru

men

t in

such

a

fash

ion

as to

gen

erat

e a

requ

ired

actu

atio

n si

gnal

.

SY-B

12

Cre

dit i

s not

take

n fo

r sys

tem

or c

ompo

nent

ope

rabi

lity

beyo

nd ra

ted

or d

esig

n ca

pabi

litie

s un

less

just

ified

.

CO

MM

ENT:

The

info

rmat

ion

whi

ch c

ould

be

used

for j

ustif

icat

ion

incl

udes

: -

test

or o

pera

tiona

l dat

a;

- ca

lcul

atio

ns;

- ve

ndor

inpu

t; -

expe

rt ju

dgm

ent.

SY-B

13

Con

tribu

tors

to sy

stem

una

vaila

bilit

y an

d un

relia

bilit

y (i.

e., c

ompo

nent

s and

spec

ific

failu

re

mod

es) m

ay b

e ex

clud

ed fr

om th

e m

odel

if o

ne o

f the

follo

win

g sc

reen

ing

crite

ria is

met

: a)

A

com

pone

nt m

ay b

e ex

clud

ed fr

om th

e sy

stem

mod

el if

the

tota

l fai

lure

pro

babi

lity

of

the

com

pone

nt fa

ilure

mod

es re

sulti

ng in

the

sam

e ef

fect

on

syst

em o

pera

tion

is a

t lea

st

two

orde

rs o

f mag

nitu

de lo

wer

than

the

high

est f

ailu

re p

roba

bilit

y of

the

othe

r co

mpo

nent

s in

the

sam

e sy

stem

trai

n th

at re

sults

in th

e sa

me

effe

ct o

n sy

stem

ope

ratio

n.

b)

One

or m

ore

failu

re m

odes

for a

com

pone

nt m

ay b

e ex

clud

ed fr

om th

e sy

stem

s mod

el if

th

e co

ntrib

utio

n of

them

to th

e to

tal f

ailu

re ra

te o

r pro

babi

lity

is le

ss th

an 1

% o

f the

tota

l fa

ilure

rate

or p

roba

bilit

y fo

r tha

t com

pone

nt, t

akin

g in

to a

ccou

nt th

e sa

me

effe

ct o

n sy

stem

ope

ratio

n, o

r

c)

The

scre

ened

con

tribu

tors

are

pos

ition

faul

ts fo

r com

pone

nts (

such

as t

hose

that

occ

ur

durin

g or

follo

win

g te

st a

nd m

aint

enan

ce a

ctiv

ities

) for

whi

ch th

e co

mpo

nent

rece

ives

an

auto

mat

ic si

gnal

to p

lace

it in

its r

equi

red

stat

e an

d no

oth

er p

ositi

on fa

ults

exi

sts (

e.g.

pu

lled

brea

kers

) tha

t wou

ld p

recl

ude

the

com

pone

nt fr

om re

ceiv

ing

the

sign

al, o

r

RA

TIO

NA

LE: O

ther

com

pone

nt fa

ilure

mod

es w

ill b

e do

min

atin

g co

ntrib

utor

s to

syst

em u

nava

ilabi

lity.

In a

dditi

on, C

CF

cont

ribut

ion

is li

kely

to b

e do

min

ated

by

com

pone

nts a

nd fa

ilure

mod

es th

at a

re

not e

xclu

ded.

57

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

d)

It

is sh

own

that

the

omis

sion

of t

he c

ontri

buto

r doe

s not

hav

e a

sign

ifica

nt im

pact

on

the

resu

lts.

Com

pone

nts o

r fai

lure

mod

es u

sing

crit

eria

(a),

(b),

or (c

) if t

hey

coul

d fa

il m

ultip

le sy

stem

s or

mul

tiple

trai

ns o

f a sy

stem

AR

E N

OT

SCR

EEN

ED.

Sp

ecia

l Attr

ibut

e SY

-B13

-S1:

Th

e ba

sic

even

t4 scre

enin

g pr

oces

s is r

evis

ited

for c

erta

in a

pplic

atio

ns.

RA

TIO

NAL

E: A

che

ckin

g w

ill b

e ne

eded

to m

ake

sure

that

or

igin

ally

scre

ened

out

eve

nts d

o no

t hav

e an

impa

ct o

n ap

plic

atio

n re

sults

. SY

-B14

N

o cr

edit

is g

iven

to re

pair

(rec

over

y) o

f har

dwar

e fa

ults

, unl

ess t

he fe

asib

ility

of r

epai

r is

just

ified

.

SY-B

15

Whe

n si

mpl

ified

mod

els,

such

as s

ingl

e ba

sic

even

t mod

ellin

g an

d gr

oupi

ng o

f bas

ic e

vent

s in

to su

per c

ompo

nent

s, ar

e us

ed, p

oten

tial s

ourc

es o

f dep

ende

ncie

s are

con

side

red

expl

icitl

y.

RA

TIO

NA

LE: S

impl

ified

mod

els m

ay m

ask

cont

ribut

ions

to th

e re

sults

of s

uppo

rt sy

stem

s or o

ther

dep

ende

nt-fa

ilure

mod

es.

EXA

MPL

ES:

Exam

ples

of d

epen

denc

ies t

hat a

re n

eede

d to

be

cons

ider

ed

expl

icitl

y in

clud

e op

erat

or a

ctio

ns, f

unct

iona

l dep

ende

ncie

s, sh

ared

co

mpo

nent

s, et

c.

Syst

ems t

hat s

omet

imes

hav

e no

t bee

n m

odel

led

in d

etai

l inc

lude

the

scra

m sy

stem

, the

pow

er-c

onve

rsio

n sy

stem

, ins

trum

ent a

ir, a

nd th

e ke

ep-f

ill s

yste

ms.

4 ‘Bas

ic e

vent

’ is a

n el

emen

t rep

rese

ntin

g an

eve

nt in

a sy

stem

faul

t mod

el th

at re

quire

s no

furth

er d

ecom

posi

tion

(e.g

. ‘pu

mp

fails

to st

art w

hen

dem

ande

d’, ‘

pum

p is

una

vaila

ble

due

to te

st o

r m

aint

enan

ce’,

etc.

).

58

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

Sp

ecia

l Attr

ibut

e SY

-B15

-S1:

Ea

ch sy

stem

is m

odel

led

with

sepa

rate

bas

ic e

vent

s dow

n to

the

leve

l of

deta

il re

quir

ed fo

r sup

port

ing

a sp

ecifi

c ap

plic

atio

n.

RATI

ON

ALE :

Cer

tain

app

licat

ions

may

requ

ire

a de

taile

d m

odel

ling

to ta

ke in

to a

ccou

nt d

iffer

ence

s bet

wee

n co

mpo

nent

s gr

oupe

d to

geth

er. E

xam

ples

of d

iffer

ence

s inc

lude

: -

hard

war

e fa

ilure

s tha

t are

not

reco

vera

ble

vers

us a

ctua

tion

sign

als w

hich

are

reco

vera

ble;

-

even

ts w

ith d

iffer

ent r

ecov

ery

pote

ntia

l; -

HE

even

ts th

at c

an h

ave

diffe

rent

pro

babi

litie

s dep

ende

nt o

n th

e co

ntex

t of d

iffer

ent a

ccid

ent s

eque

nces

; -

even

ts w

hich

are

mut

ually

exc

lusi

ve o

f oth

er e

vent

s not

in th

e m

odul

e;

- ev

ents

whi

ch o

ccur

in o

ther

faul

t tre

es (e

spec

ially

com

mon

-ca

use

even

ts);

-

com

pone

nts h

avin

g di

ffere

nt m

aint

enan

ce a

nd te

stin

g st

rate

gies

; -

SSC

s use

d by

oth

er sy

stem

s.

SY-B

16

An

appr

opria

te re

liabi

lity

mod

el, t

hat m

atch

es th

e de

finiti

ons a

nd d

ata

avai

labl

e, is

use

d fo

r ea

ch b

asic

eve

nt.

EXA

MPL

ES: R

elia

bilit

y m

odel

s in

use

in d

iffer

ent P

SAs i

nclu

de:

1. M

onito

red,

repa

irabl

e (s

tand

by fa

ilure

rate

, rep

air t

ime)

2.

Per

iodi

cally

test

ed (

stan

dby

failu

re ra

te, t

est i

nter

val)

3. C

onst

ant u

nava

ilabi

lity

(pro

babi

lity)

4.

Fix

ed m

issi

on ti

me

(ope

ratin

g fa

ilure

rate

, mis

sion

tim

e)

5. N

on re

paira

ble

(sta

ndby

failu

re ra

te)

Sp

ecia

l Attr

ibut

e SY

-B16

-S1:

Ti

me-

depe

nden

t fai

lure

mod

els a

re u

sed.

RA

TIO

NAL

E : C

erta

in a

pplic

atio

ns, e

g te

st in

terv

al o

ptim

isat

ion,

re

quir

e th

e us

e of

tim

e de

pend

ent m

odel

s.

SY-B

17

The

even

t nam

ing

sche

me

is d

evel

oped

in a

con

sist

ent m

anne

r (se

e al

so T

able

9.2

-A, G

ener

al

Attr

ibut

e D

A-A

02).

RA

TIO

NA

LE: M

odel

man

ipul

atio

n an

d in

terp

reta

tion

is fa

cilit

ated

by

for e

xam

ple

usin

g th

e sa

me

desi

gnat

or fo

r a c

ompo

nent

type

and

fa

ilure

mod

e.

59

T

able

7.2

-C

Att

ribu

tes f

or S

yste

ms A

naly

sis:

Tas

k SY

-C ‘I

dent

ifica

tion

and

Mod

ellin

g of

Dep

ende

ncie

s’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

SY-C

A

ll de

pend

enci

es a

re id

entif

ied.

Des

ign

rela

ted

and

oper

atio

nal d

epen

denc

ies,

both

di

rect

and

indi

rect

, are

exp

licitl

y m

odel

led

as fa

r as p

ossi

ble.

Res

idua

l dep

ende

ncie

s ar

e ac

coun

ted

for b

y C

CF

mod

ellin

g.

All

com

pone

nts o

r fai

lure

mod

es, w

hich

cou

ld fa

il m

ultip

le s

yste

ms (

so c

alle

d sh

ared

co

mpo

nent

s), a

re in

clud

ed in

the

mod

el, e

ven

if th

e in

depe

nden

t im

pact

of t

he

com

pone

nt/fa

ilure

mod

e is

con

side

red

non-

sign

ifica

nt.

RA

TIO

NA

LE: F

ailu

re o

f sha

red

com

pone

nts m

ay h

ave

a si

gnifi

cant

co

ntrib

utio

n to

risk

.

EXA

MPL

E: C

omm

on su

ctio

n pi

pe fe

edin

g tw

o sy

stem

s.

SY-C

01

Spec

ial A

ttrib

ute

SY

-C01

-S1

Whe

n pi

pe ru

ptur

e/le

aks a

re in

clud

ed in

the

mod

el, t

he e

ffect

of p

ipe

failu

re o

n th

e ef

fect

iven

ess o

f all

conn

ecte

d co

mpo

nent

s (e.

g.

pum

ps, h

eat e

xcha

nger

s) is

mod

elle

d.

RATI

ON

ALE:

PSA

app

licat

ions

dea

ling

with

opt

imiz

atio

n of

ISI (

i.e. r

isk-

info

rmed

ISI)

requ

ire

adeq

uate

mod

ellin

g of

the

impa

ct o

f pip

e ru

ptur

es if

th

e la

tter a

re in

clud

ed in

the

mod

el.

EXAM

PLE:

Whe

n tw

o pu

mps

are

con

nect

ed to

the

sam

e pi

pe, a

rupt

ure

of

this

pip

e w

ould

dis

able

bot

h pu

mps

. Thi

s sho

uld

be c

orre

ctly

acc

ount

ed fo

r in

the

mod

el o

f eac

h pu

mp.

SY-C

02

A su

bcom

pone

nt th

at is

shar

ed b

y m

ore

than

one

com

pone

nt o

r affe

cts a

noth

er

com

pone

nt is

mod

elle

d se

para

tely

.

SY-C

03

Supp

ort f

unct

ions

and

syst

ems n

eede

d to

per

form

the

syst

em m

issi

on(s

) are

mod

elle

d ex

plic

itly.

Just

ifica

tion

is p

rovi

ded

for c

ases

whe

re d

epen

denc

ies a

re e

xclu

ded

from

the

mod

el.

Cau

tion:

A c

aref

ul id

entif

icat

ion

(map

ping

) and

mod

ellin

g of

con

trol l

ogic

and

pow

er

depe

nden

cies

is n

eede

d to

acc

ount

for a

ll de

pend

enci

es in

this

are

a.

EXA

MPL

E of

supp

ort f

unct

ions

: -

Act

uatio

n lo

gic

incl

udin

g pr

esen

ce o

f con

ditio

ns n

eede

d fo

r aut

omat

ic

actu

atio

n an

d pe

rmis

sive

and

lock

out s

igna

ls

- Su

ppor

t sys

tem

s req

uire

d fo

r con

trol o

f com

pone

nts

- C

ompo

nent

mot

ive

pow

er

- C

oolin

g of

com

pone

nts

- Sc

reen

cle

anin

g sy

stem

-

Hea

ting/

vent

ilatio

n sy

stem

-

Wat

er m

ake-

up

- O

verp

ress

ure

prot

ectio

n -

Any

oth

er id

entif

ied

supp

ort f

unct

ion

nece

ssar

y to

mee

t the

succ

ess

crite

ria a

nd a

ssoc

iate

d sy

stem

s

SY-C

04

Mis

sion

tim

es fo

r sup

port

syst

ems a

re m

odel

led

cons

iste

nt w

ith th

e m

issi

on ti

me

for

the

fron

t lin

e sy

stem

s and

in a

ccor

danc

e w

ith th

e su

cces

s crit

eria

(see

Tas

k SC

-B in

Se

ctio

n 6)

.

60

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

SY-C

05

The

avai

labl

e in

vent

orie

s of f

uel,

wat

er in

tank

s etc

are

com

pare

d w

ith th

ose

requ

ired

to su

ppor

t eac

h su

cces

s crit

erio

n an

d th

e m

odel

refle

cts t

he re

sults

of t

his c

ompa

rison

. R

ATI

ON

ALE

: Diff

eren

t acc

iden

t sce

nario

s may

requ

ire d

iffer

ent

inve

ntor

ies.

In so

me

case

s inv

ento

ry is

insu

ffic

ient

to c

ompl

ete

the

mis

sion

an

d su

pple

men

tary

sour

ces m

ay b

e re

quire

d. S

uch

sour

ces w

ill n

eed

to b

e in

clud

ed in

the

syst

em m

odel

.

EXA

MPL

E of

inve

ntor

ies:

-

Acc

umul

ator

air

inve

ntor

y -

Bat

tery

life

-

Die

sel f

uel t

ank

capa

city

-

Wat

er st

orag

e ta

nks

SY-C

06

The

mod

ellin

g of

ele

ctric

pow

er su

pply

con

side

rs d

iffer

ent p

ower

supp

ort c

ases

to

acco

unt f

or b

atte

ry p

ower

ava

ilabi

lity

until

rest

orat

ion

of a

uxili

ary

and

exte

rnal

pow

er.

EXA

MPL

ES o

f diff

eren

t pow

er su

pply

cas

es a

re:

- A

ll po

wer

sour

ces c

an b

e ac

coun

ted

for i

nitia

lly

- B

atte

ry p

ower

ava

ilabl

e in

itial

ly (b

atte

ry) f

or c

onne

ctin

g to

aux

iliar

y po

wer

-

Long

term

cas

e w

ithou

t bat

tery

ava

ilabi

lity,

but

oth

er p

ower

sour

ce is

re

stor

ed.

CO

MM

ENT :

Bat

terie

s can

fail

to ta

ke lo

ad w

hen

dem

ande

d. T

his f

ailu

re

mod

e ne

eds t

o be

con

side

red.

The

pro

babi

lity

of su

ch e

vent

dep

ends

on

desi

gn a

nd o

pera

tiona

l fea

ture

s of t

he b

atte

ry a

nd b

atte

ry c

harg

ing

syst

em.

SY-C

07

Syst

em c

ondi

tions

that

cau

se a

loss

of d

esire

d sy

stem

func

tion

incl

udin

g co

nditi

ons

that

cau

se th

e sy

stem

to is

olat

e or

trip

, are

mod

elle

d ex

plic

itly

by u

sing

real

istic

fu

nctio

nal r

equi

rem

ents

that

are

supp

orte

d by

eng

inee

ring

anal

ysis

.

If e

ngin

eerin

g an

alys

es a

nd d

ata

are

not a

vaila

ble

to ju

stify

cer

tain

failu

re p

roba

bilit

y,

the

equi

pmen

t/sys

tem

failu

re p

roba

bilit

y is

ass

umed

to b

e 1.

0.

EXA

MPL

ES o

f con

ditio

ns th

at is

olat

e or

trip

a sy

stem

incl

ude:

-

Syst

em-r

elat

ed p

aram

eter

s suc

h as

a h

igh

tem

pera

ture

with

in th

e sy

stem

-

Exte

rnal

par

amet

ers u

sed

to p

rote

ct th

e sy

stem

from

oth

er fa

ilure

s (e.

g.

the

high

reac

tor p

ress

ure

vess

el (R

PV) w

ater

leve

l iso

latio

n si

gnal

use

d to

pre

vent

wat

er in

trusi

on in

to th

e tu

rbin

es o

f the

RC

IC a

nd H

PCI

pum

ps o

f a B

WR

) -

Adv

erse

env

ironm

enta

l con

ditio

ns.

CO

MM

ENT:

Equ

ipm

ent p

rote

ctio

n si

gnal

s may

cau

se a

dire

ct a

utom

atic

tri

p fo

r som

e co

mpo

nent

s, or

equ

ipm

ent p

roce

dure

s req

uire

a m

anua

l trip

. B

oth

way

s nee

d to

be

cons

ider

ed if

app

licab

le.

Sp

ecia

l Attr

ibut

e SY

-C07

-S1:

En

gine

erin

g an

alys

is is

use

d to

just

ify c

erta

in fa

ilure

pro

babi

lity

due

to sp

ecifi

c sy

stem

con

ditio

ns.

RATI

ON

ALE :

Con

serv

ativ

e ap

proa

ch m

ay n

ot b

e po

ssib

le in

an

adva

nced

ap

plic

atio

n, w

here

it c

an h

ide

impo

rtan

t res

ults

.

61

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

SY-C

08

The

syst

ems p

oten

tial f

or c

ausi

ng a

CC

I is c

onsi

dere

d (s

ee a

lso

Tabl

e 4.

2-E,

Gen

eral

A

ttrib

ute

IE-E

03).

SY-C

09

SSC

s tha

t may

be

requ

ired

to o

pera

te in

con

ditio

ns b

eyon

d th

eir e

nviro

nmen

tal

qual

ifica

tions

are

iden

tifie

d an

d de

pend

ent f

ailu

res o

f mul

tiple

SSC

s tha

t res

ult f

rom

op

erat

ion

in th

ese

adve

rse

cond

ition

s are

incl

uded

.

EXA

MPL

ES o

f deg

rade

d en

viro

nmen

ts in

clud

e:

- LO

CA

insi

de c

onta

inm

ent w

ith fa

ilure

of c

onta

inm

ent h

eat r

emov

al;

- sa

fety

relie

f val

ve (i

n dr

ywel

l) op

erab

ility

in c

ase

of sm

all L

OC

A, w

ith

dryw

ell s

pray

in a

BW

R;

- hi

gh e

nerg

y lin

e br

eaks

, e. g

., st

eam

line

bre

aks o

utsi

de c

onta

inm

ent;

- de

bris

that

cou

ld p

lug

scre

ens/

filte

rs (b

oth

inte

rnal

and

ext

erna

l to

the

plan

t);

- he

atin

g of

the

wat

er su

pply

(e.g

. BW

R su

ppre

ssio

n po

ol, P

WR

co

ntai

nmen

t sum

p) th

at c

ould

affe

ct p

ump

oper

abili

ty;

- st

eam

bin

ding

of p

umps

; -

cont

ainm

ent v

ent a

nd fa

ilure

eff

ects

.

SY-C

10

The

loca

tions

of c

ompo

nent

s vul

nera

ble

to e

nviro

nmen

tal h

azar

ds th

at m

ay im

pact

sy

stem

ope

ratio

n ar

e id

entif

ied.

SY-C

11

Spat

ial a

nd e

nviro

nmen

tal h

azar

ds re

sulti

ng fr

om th

e ac

cide

nt se

quen

ce sc

enar

ios

stud

ied

that

may

impa

ct s

yste

m o

pera

tion

are

iden

tifie

d an

d ac

coun

ted

for i

n th

e sy

stem

faul

t tre

e or

the

acci

dent

sequ

ence

eva

luat

ion.

EXA

MPL

E: U

se re

sults

of p

lant

wal

k do

wns

as a

sour

ce o

f inf

orm

atio

n an

d re

solu

tion

of is

sues

in th

e ev

alua

tion

of th

eir i

mpa

cts.

SY-C

12

Ope

rato

r int

erfa

ce d

epen

denc

ies a

cros

s sys

tem

s or t

rain

s are

con

side

red

whe

re

appl

icab

le. (

See

also

Tab

le 8

.2-A

, Gen

eral

Attr

ibut

e H

R-A

02, T

able

8.2

-B, G

ener

al

Attr

ibut

e H

R-B

01, T

able

8.2

-D, G

ener

al A

ttrib

ute

HR

-D04

, and

Tab

le 8

.2-G

, Gen

eral

A

ttrib

ute

HR

-G07

).

RA

TIO

NA

LE: S

ever

al o

pera

tor a

ctio

ns re

lyin

g on

the

sam

e in

terf

ace

have

to

be

treat

ed a

s non

-inde

pend

ent e

vent

s in

orde

r to

avoi

d to

o op

timis

tic

resu

lts.

SY-C

13

Intra

-sys

tem

com

mon

cau

se fa

ilure

s are

mod

elle

d us

ing

an a

ccep

tabl

e m

odel

ling

appr

oach

.

RA

TIO

NA

LE: C

omm

on c

ause

failu

res a

re d

omin

atin

g co

ntrib

utor

to ri

sk in

pl

ants

rely

ing

on re

dund

anci

es a

s a m

eans

of a

chie

ving

a h

igh

relia

bilit

y.

EXA

MPL

E: E

xam

ples

of m

etho

ds a

re a

vaila

ble

in R

ef. [

13] a

nd R

ef. [

5].

SY-C

14

Inte

r-sy

stem

com

mon

cau

se fa

ilure

s are

con

side

red

for c

ompo

nent

s in

syst

ems t

hat a

re

shar

ed b

etw

een

diff

eren

t pla

nts.

EX

AM

PLE:

Mul

tiple

uni

t cro

sstie

s bet

wee

n th

e D

Gs o

f tw

o un

its a

t the

sa

me

site

.

62

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

Sp

ecia

l Attr

ibut

e SY

-C14

-S1:

In

ter-

syst

em c

omm

on c

ause

failu

res a

re c

onsi

dere

d.

RATI

ON

ALE:

Con

side

ratio

n of

inte

r-sy

stem

com

mon

cau

se fa

ilure

s may

be

need

ed to

avo

id u

njus

tifie

d op

timis

tic re

sults

.

EXAM

PLE:

The

sam

e ty

pe o

f val

ve a

cros

s a n

umbe

r of s

yste

ms.

CO

MM

ENT:

For

the

plan

ts w

ith h

igh

redu

ndan

cy in

term

s of s

yste

ms

perf

orm

ing

the

sam

e sa

fety

func

tion

inte

r-sy

stem

CC

F m

ay b

e th

e m

ost

impo

rtan

t con

trib

utor

to th

e fu

nctio

n fa

ilure

.

SY-C

15

Com

mon

cau

se c

ompo

nent

gro

ups (

CC

CG

s) a

re d

efin

ed b

ased

on

a lo

gica

l, sy

stem

atic

ju

stifi

ed p

roce

ss th

at c

onsi

ders

sim

ilarit

y.

CO

MM

ENTS

:

- Ty

pica

lly th

is su

gges

ts th

at C

CC

Gs a

re d

efin

ed w

ithin

the

sam

e si

ngle

sy

stem

at t

he sa

me

plan

t, bu

t whe

n tw

o or

mor

e sy

stem

s are

ess

entia

lly

iden

tical

and

ope

rate

d in

the

sam

e m

anne

r the

y ca

n be

con

side

red

acro

ss sy

stem

bou

ndar

ies.

- C

onsi

dera

tion

of d

iffer

ent f

unct

iona

l fai

lure

mod

es m

ay n

eed

defin

ition

of

mor

e th

an o

ne C

CC

G fo

r the

sam

e se

t of c

ompo

nent

s. A

n ex

ampl

e is

th

e ca

se o

f saf

ety/

relie

f val

ves

whe

re th

e fa

ilure

to o

pen

and

failu

re to

re

-clo

se a

fter o

peni

ng a

re tr

eate

d by

two

CC

CG

s, on

e fo

r eac

h fa

ilure

m

ode.

EXA

MPL

E: T

ypic

al si

mila

ritie

s use

d fo

r the

def

initi

on o

f com

mon

cau

se

com

pone

nt g

roup

s are

: des

ign/

hard

war

e, fu

nctio

n, in

stal

latio

n,

mai

nten

ance

, tes

t int

erva

l, pr

oced

ures

, ser

vice

con

ditio

ns, l

ocat

ion

and

envi

ronm

ent,

man

ufac

ture

r. C

andi

date

s for

com

mon

-cau

se fa

ilure

gro

ups

incl

ude

both

act

ive

and

pass

ive

com

pone

nts s

uch

as: m

otor

-ope

rate

d va

lves

, pu

mps

, saf

ety-

relie

f val

ves,

air-

oper

ated

val

ves,

sole

noid

-ope

rate

d va

lves

, ch

eck

valv

es, d

iese

l gen

erat

ors,

batte

ries,

inve

rters

and

bat

tery

cha

rger

, ci

rcui

t bre

aker

s, sc

ram

val

ves,

RPS

logi

c ch

anne

ls, R

PS lo

gic

sens

ors,

rela

ys, s

train

ers,

elec

trica

l bus

es, c

hille

rs, c

ompr

esso

rs, c

ontro

l rod

s, ai

r dr

yers

, fus

es (w

ire a

nd e

lect

roni

c), e

lect

ric h

eate

rs, s

witc

hes,

trans

form

ers,

vent

ilatio

n fla

ps, v

entil

atio

n fa

ns, e

tc.

Sp

ecia

l Attr

ibut

e SY

-C15

-S1:

D

iver

sifie

d co

mpo

nent

s can

nor

mal

ly b

e co

nsid

ered

to b

e in

depe

nden

t. H

owev

er, i

f the

div

ersi

fied

com

pone

nts h

ave

iden

tical

pa

rts,

ther

e m

ay b

e a

need

to b

reak

dow

n th

e co

mpo

nent

s int

o sm

alle

r par

ts, a

nd m

odel

iden

tical

par

ts a

s CC

CG

s.

RATI

ON

ALE:

Cer

tain

app

licat

ion

requ

irin

g a

high

er le

vel o

f det

ail m

ay

also

requ

ire

a hi

gher

leve

l of d

etai

l in

CC

CG

def

initi

ons.

63

T

able

7.2

-D

Att

ribu

tes f

or S

yste

ms A

naly

sis:

Tas

k SY

-D ‘D

ocum

enta

tion’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

SY-D

D

ocum

enta

tion

and

info

rmat

ion

stor

age

is p

erfo

rmed

in a

man

ner t

hat f

acili

tate

s pee

r re

view

, as w

ell a

s fut

ure

upgr

ades

and

app

licat

ions

of t

he P

SA b

y de

scrib

ing

the

proc

esse

s th

at w

ere

used

and

pro

vidi

ng d

etai

ls o

f the

ass

umpt

ions

mad

e an

d th

eir b

ases

.

SY-D

01

The

proc

esse

s tha

t wer

e fo

llow

ed to

sele

ct, t

o m

odel

, and

to q

uant

ify th

e sy

stem

un

avai

labi

lity

are

docu

men

ted.

Ass

umpt

ions

and

bas

es a

re st

ated

SY-D

02

The

follo

win

g is

doc

umen

ted:

-

Rev

isio

n hi

stor

y -

Ope

n is

sues

(que

stio

ns) a

nd a

nsw

ers o

n pr

evio

us is

sues

-

The

codi

ng sy

stem

use

d fo

r the

PSA

and

PSA

mod

el

- Sy

stem

func

tion

and

oper

atio

n un

der n

orm

al a

nd a

ccid

ent c

ondi

tions

-

Syst

em a

ctiv

atio

n/bl

ocki

ng

- A

ltern

ativ

e sy

stem

alig

nmen

ts

- Sy

stem

bou

ndar

y -

Syst

em sc

hem

atic

illu

stra

ting

all e

quip

men

t and

com

pone

nts n

eces

sary

for s

yste

m

oper

atio

n an

d co

mpo

nent

s tha

t are

mod

elle

d -

Dep

ende

ncy

mat

rices

on

com

pone

nt le

vel (

it is

use

ful t

o pr

oduc

e an

inte

grat

ed

depe

nden

cy m

atrix

to p

rovi

de a

n ov

ervi

ew o

f the

func

tiona

l dep

ende

ncie

s ove

r all

syst

ems)

-

Info

rmat

ion

and

calc

ulat

ions

to su

ppor

t equ

ipm

ent o

pera

bilit

y co

nsid

erat

ions

and

as

sum

ptio

ns

- A

ctua

l ope

ratio

nal h

isto

ry in

dica

ting

any

past

pro

blem

s in

the

syst

em o

pera

tion

mod

ifica

tion

hist

ory

- Sy

stem

succ

ess c

riter

ia a

nd re

latio

nshi

p to

acc

iden

t seq

uenc

e m

odel

s -

Hum

an a

ctio

ns n

eces

sary

for o

pera

tion

of sy

stem

-

Ref

eren

ce to

syst

em-r

elat

ed te

st a

nd m

aint

enan

ce p

roce

dure

s -

Syst

em d

epen

denc

ies a

nd sh

ared

com

pone

nt in

terf

ace

- C

ompo

nent

bou

ndar

ies

- C

ompo

nent

spat

ial i

nfor

mat

ion

64

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

- A

ssum

ptio

ns o

r sim

plifi

catio

ns m

ade

in d

evel

opm

ent o

f the

syst

em m

odel

s -

The

syst

ems p

oten

tial f

or b

eing

a C

omm

on C

ause

Initi

ator

-

A li

st o

f all

com

pone

nts a

nd fa

ilure

mod

es in

clud

ed in

the

mod

el (t

he b

asic

eve

nts)

al

ong

with

just

ifica

tion

for a

ny e

xclu

sion

of c

ompo

nent

s and

failu

re m

odes

-

A li

st o

f all

hum

an a

ctio

n fa

ilure

mod

es in

clud

ed in

the

mod

el

- A

list

of a

ll C

CC

Gs i

nclu

ded

in th

e m

odel

-

A d

escr

iptio

n of

the

mod

ular

izat

ion

proc

ess (

if us

ed)

- R

ecor

ds o

f res

olut

ion

of lo

gic

loop

s dev

elop

ed d

urin

g fa

ult t

ree

linki

ng (i

f use

d)

- R

esul

ts o

f the

sys

tem

mod

el e

valu

atio

ns

- R

esul

ts o

f sen

sitiv

ity st

udie

s (if

used

) -

The

sour

ces o

f the

abo

ve in

form

atio

n, (e

.g. c

ompl

eted

che

cklis

t fro

m w

alk

dow

ns,

note

s fro

m d

iscu

ssio

ns w

ith p

lant

per

sonn

el)

- Fa

ult t

ree

desc

riptio

n in

clud

ing

cond

ition

s for

the

mod

el, t

op g

ate,

faul

t tre

e la

yout

, tra

nsfe

rs, h

ouse

eve

nts,

attri

bute

s, fa

ilure

mod

es, C

CF

mod

ellin

g, te

st a

nd

mai

nten

ance

mod

ellin

g, o

pera

tor a

ctio

ns a

nd si

gnal

mod

ellin

g.

SY-D

03

The

inpu

t and

resu

lts o

f any

scre

enin

g pr

oces

ses s

hall

be st

ored

for f

utur

e re

-ana

lysi

s, w

hich

may

be

nece

ssar

y fo

r app

licat

ions

or u

pdat

e of

the

PSA

.

SY-D

04

Stor

age

of th

e sy

stem

s ana

lysi

s inf

orm

atio

n.

The

info

rmat

ion

from

the

syst

ems a

naly

sis (

e g

FMEA

and

faul

t tre

es w

ith g

ates

and

bas

ic

even

ts) i

s par

t of t

he P

SA m

odel

. The

PSA

mod

el is

stor

ed a

nd d

etai

led

back

grou

nd

info

rmat

ion

is st

ored

in a

retri

evab

le a

nd a

cces

sibl

e el

ectro

nic

form

and

form

at.

65

8. PSA ELEMENT ‘HR’: HUMAN RELIABILITY ANALYSIS


The objective of the human reliability analysis is to incorporate in the PSA model the impact of plant personnel actions on risk. The personnel actions considered are of two types: the first type includes those associated with the performance of surveillance testing, maintenance, and calibration, often referred to as pre-initiating event actions; the second type includes those associated with responses to plant disturbances as outlined in emergency and off-normal operating procedures or their equivalent, often referred to as post-initiating event actions. When constructing the PSA model these translate into the assessment of what in Safety Series No. 50-P-10 (Ref. [6]) are referred to as Type A and Type C human action events. When constructing models to estimate frequencies for the support system initiating events (Type B human action events), both types of actions are considered.

The important HRA topics are:

- The identification of the specific human activities whose impact should be included in the analysis;

- The representation of the impact of success or failure to perform those activities correctly in the accident sequence models (e.g. event trees) and the supporting system reliability models (e.g. fault trees);

- The estimation of the probabilities of the logic model events (sometimes called human failure events [HFEs] or human errors [HE]) representing the contribution of the operators’ failure to perform the required actions correctly as specific modes of unavailability of the component, system, or function affected.

The analyses on the above topics should be performed using a systematic process and

in such a way that the plant-specific and, where necessary, accident scenario specific factors are taken into account. In addition, the dependency between the human failure events should be properly characterized and taken into account to ensure that the accident sequence frequency estimations are performed correctly. 8.2. Human reliability analysis tasks and their attributes

Table 8.2 lists the main tasks for the PSA element ‘Human Reliability Analysis’.

Tables 8.2-A through 8.2-I present the description of general and special attributes for these tasks.

66

Table 8.2 Main Tasks for Human Reliability Analysis


Pre-initiating event HRA

HR-A Identification of Routine Activities

HR-B Screening of Activities

HR-C Definition of Pre-initiator Human Failure Events

HR-D Assessment of Probabilities of Pre-initiator Human Failure Events

Post-initiating event HRA

HR-E Identification of Post-Initiator Operator Responses

HR-F Definition of Post-initiator Human Failure Events

HR-G Assessment of Probabilities of Post-initiator Human Failure Events

HR-H Recovery Actions

Documentation

HR-I Documentation

67

T

able

8.2

-A

Att

ribu

tes f

or H

uman

Rel

iabi

lity

Ana

lysi

s: T

ask

HR

-A ‘I

dent

ifica

tion

of R

outin

e A

ctiv

ities

(Pre

-initi

atin

g E

vent

HR

A)’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

HR

-A

The

task

incl

udes

a th

roug

h id

entif

icat

ion

of sp

ecifi

c ro

utin

e ac

tiviti

es, w

hich

, if n

ot

perf

orm

ed c

orre

ctly

, im

pact

the

avai

labi

lity

of e

quip

men

t nec

essa

ry to

per

form

the

syst

em

func

tions

mod

elle

d in

the

PSA

.

HR

-A01

Th

roug

h a

revi

ew o

f pro

cedu

res a

nd o

pera

tiona

l pra

ctic

es, t

hose

test

and

mai

nten

ance

ac

tiviti

es th

at re

quire

real

ignm

ent o

f equ

ipm

ent f

rom

its n

orm

al o

pera

tiona

l or s

tand

by

stat

us a

re id

entif

ied

for t

hose

sys

tem

s and

com

pone

nts r

equi

red

to p

erfo

rm th

e fu

nctio

ns

requ

ired

to re

spon

d to

the

initi

atin

g ev

ents

mod

elle

d.

CO

MM

ENT:

The

inte

nt is

to id

entif

y th

ose

oppo

rtuni

ties f

or o

pera

tors

to

fail

to re

alig

n eq

uipm

ent t

o its

requ

ired

stat

us fo

llow

ing

com

plet

ion

of th

e ac

tivity

, suc

h th

at it

wou

ld b

e un

avai

labl

e sh

ould

it b

e ca

lled

upon

to

resp

ond

to a

n in

itiat

ing

even

t. P

artic

ular

atte

ntio

n sh

ould

be

paid

to

activ

ities

that

can

dis

able

mul

tiple

trai

ns o

f a sy

stem

sim

ulta

neou

sly

(e.g

. th

e au

tom

atic

initi

atio

n of

the

stan

dby

liqui

d co

ntro

l sys

tem

in a

BW

R is

ty

pica

lly d

isab

led

for t

est p

urpo

ses)

. It

is ty

pica

lly a

ssum

ed th

at th

e co

ntrib

utio

n to

com

pone

nt u

nava

ilabi

lity

from

failu

res r

esul

ting

from

in

corr

ectly

per

form

ed m

aint

enan

ce a

re c

aptu

red

in th

e eq

uipm

ent f

ailu

re

prob

abili

ty.

Thus

, the

focu

s her

e is

on

the

failu

re to

real

ign

equi

pmen

t up

on c

ompl

etio

n of

mai

nten

ance

.

HR

-A02

Th

roug

h a

revi

ew o

f pro

cedu

res a

nd p

ract

ices

, tho

se c

alib

ratio

n ac

tiviti

es a

re id

entif

ied

that

, if p

erfo

rmed

inco

rrec

tly, h

ave

the

capa

bilit

y of

def

eatin

g th

e au

tom

atic

initi

atio

n of

st

andb

y sa

fety

equ

ipm

ent o

r of r

ende

ring

requ

ired

func

tions

of s

yste

ms o

r com

pone

nts

unav

aila

ble.

In p

artic

ular

, tho

se a

ctiv

ities

that

hav

e th

e po

tent

ial t

o af

fect

equ

ipm

ent i

n m

ultip

le tr

ains

of

a re

dund

ant s

yste

m, o

r in

dive

rse

syst

ems,

e.g.

as a

resu

lt of

usi

ng in

appr

opria

te c

alib

ratio

n pr

oced

ures

or f

aulty

or i

mpr

oper

ly c

alib

rate

d ca

libra

tion

equi

pmen

t, ar

e id

entif

ied.

The

proc

edur

es w

hich

allo

w d

etec

ting

the

faul

ty a

lignm

ent o

r cal

ibra

tion

are

also

revi

ewed

.

EXA

MPL

ES: i

ncor

rect

cal

ibra

tion

of st

eam

gen

erat

or le

vel s

enso

rs;

inco

rrec

t set

ting

of to

rque

sw

itche

s.

68

T

able

8.2

-B

Att

ribu

tes f

or H

uman

Rel

iabi

lity

Ana

lysi

s: T

ask

HR

-B ‘S

cree

ning

of A

ctiv

ities

(Pre

-initi

atin

g E

vent

HR

A)’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

HR

-B

The

task

is a

imed

at s

cree

ning

of a

ctiv

ities

that

, on

the

basi

s of t

he p

lant

pra

ctic

es, c

an b

e ar

gued

to re

sult

in a

suffi

cien

tly lo

w li

kelih

ood

of fa

ilure

. C

OM

MEN

T: T

his s

ectio

n re

fers

to th

e sc

reen

ing

of a

ctiv

ities

so th

at th

e po

tent

ial h

uman

failu

res a

ssoc

iate

d w

ith th

em a

re n

ot in

clud

ed in

the

PRA

m

odel

.

HR

-B01

C

lass

es o

f act

iviti

es th

at a

re p

erfo

rmed

in a

sim

ilar m

anne

r (e.

g. te

sts o

f MO

Vs)

are

sc

reen

ed o

nly

if th

e pl

ant p

ract

ices

are

such

that

they

can

be

argu

ed to

resu

lt in

the

prob

abili

ty o

f equ

ipm

ent n

ot b

eing

rest

ored

to st

andb

y st

atus

is sm

all c

ompa

red

with

ot

her m

odes

of u

nava

ilabi

lity

of th

at e

quip

men

t, i.e

., fa

ilure

s or u

nava

ilabi

lity

due

to

bein

g ou

t of s

ervi

ce fo

r mai

nten

ance

.

Act

iviti

es th

at re

sult

in m

ultip

le tr

ains

of a

syst

em b

eing

mad

e un

avai

labl

e du

ring

the

cour

se o

f the

act

ivity

are

ana

lyse

d in

mor

e de

tail,

sinc

e th

e pr

obab

ilitie

s are

com

pare

d to

co

mm

on c

ause

failu

re p

roba

bilit

ies.

RA

TIO

NA

LE:

It is

pos

sibl

e to

redu

ce th

e nu

mbe

r of c

ontri

buto

rs to

eq

uipm

ent u

nava

ilabi

lity

due

to h

uman

err

or to

avo

id u

nnec

essa

rily

com

plic

atin

g th

e sy

stem

mod

els,

whe

n th

ose

cont

ribut

ors d

o no

t im

pact

the

resu

lts si

gnifi

cant

ly.

EXA

MPL

E: T

he h

uman

err

or to

leav

e a

valv

e in

wro

ng p

ositi

on a

fter t

est

may

be

excl

uded

if th

ere

is a

n in

dica

tion

of th

e va

lve

posi

tion

in th

e m

ain

cont

rol r

oom

and

an

auto

mat

ic si

gnal

to re

stor

e th

e op

erat

ing

stat

us o

f the

va

lve

is su

pplie

d in

cas

e of

dem

and

for t

he sy

stem

act

uatio

n.

CO

MM

ENT:

Pro

babi

litie

s use

d fo

r scr

eeni

ng m

ay b

e ge

nera

ted

by

appl

icat

ion

of si

mpl

e H

RA

mod

els,

such

as T

HER

P (s

ee R

ef. [

14])

. (Se

e al

so T

able

8.2

-D, G

ener

al A

ttrib

ute

HR

-D02

).

HR

-B02

Fo

r uni

que,

one

-of-a

-kin

d ac

tiviti

es, s

cree

ning

is p

erfo

rmed

onl

y w

hen

it ca

n be

show

n th

at, o

n th

e ba

sis o

f the

act

ivity

spec

ific

proc

edur

es, t

he d

efen

ces i

n pl

ace

to p

reve

nt th

e fa

ilure

to re

turn

equ

ipm

ent t

o its

requ

ired

conf

igur

atio

n ar

e su

ffic

ient

to re

duce

the

failu

re

prob

abili

ty b

elow

that

of o

ther

mod

es o

f una

vaila

bilit

y of

the

equi

pmen

t.

CO

MM

ENT:

Use

of t

hese

scre

enin

g ru

les i

s typ

ical

ly ju

stifi

ed th

roug

h ap

plic

atio

n of

sim

ple

HR

A m

odel

s, su

ch a

s TH

ERP

(see

Ref

. [14

]). (

See

also

Tab

le 8

.2-D

, Gen

eral

Attr

ibut

e H

R-D

02).

EXA

MPL

ES o

f def

ence

s inc

lude

: -

Equi

pmen

t is a

utom

atic

ally

re-a

ligne

d on

syst

em d

eman

d an

d th

e as

soci

ated

con

trol c

ircui

try is

not

dis

able

d as

par

t of t

he a

ctiv

ity.

- A

full

func

tiona

l tes

t is a

lway

s per

form

ed o

n co

mpl

etio

n of

m

aint

enan

ce.

- Eq

uipm

ent s

tatu

s is i

ndic

ated

in th

e co

ntro

l roo

m, i

ts st

atus

is

mon

itore

d ro

utin

ely,

and

real

ignm

ent c

an b

e ef

fect

ed fr

om th

e co

ntro

l ro

om.

- Eq

uipm

ent s

tatu

s is r

equi

red

to b

e ch

ecke

d lo

cally

on

a fr

eque

nt b

asis

(e

.g. o

nce

a sh

ift),

and

the

indi

catio

ns o

f mis

alig

nmen

t are

cle

ar.

HR

-B03

Sc

reen

ing

of sp

ecifi

c ca

libra

tion

activ

ities

is p

erfo

rmed

onl

y w

hen

it ca

n be

show

n th

at,

on th

e ba

sis o

f pra

ctic

es a

nd p

roce

dure

s for

cal

ibra

tion,

the

likel

ihoo

d of

mis

calib

ratio

n th

at is

sign

ifica

nt e

noug

h to

dis

able

an

impo

rtant

func

tion

is sm

all i

n co

mpa

rison

with

ot

her f

ailu

res o

r mod

es o

f una

vaila

bilit

y, in

clud

ing

CC

F.

RA

TIO

NA

LE:

it is

pos

sibl

e to

redu

ce th

e nu

mbe

r of c

ontri

buto

rs to

eq

uipm

ent u

nava

ilabi

lity

to a

void

unn

eces

saril

y co

mpl

icat

ing

the

syst

em

mod

els,

whe

n th

ose

cont

ribut

ors d

o no

t im

pact

the

resu

lts si

gnifi

cant

ly.

CO

MM

ENT:

Pro

babi

litie

s use

d fo

r scr

eeni

ng m

ay b

e ge

nera

ted

by

appl

icat

ion

of si

mpl

e H

RA

mod

els,

such

as T

HER

P (s

ee R

ef. [

14])

.

69

T

able

8.2

-C

Att

ribu

tes f

or H

uman

Rel

iabi

lity

Ana

lysi

s: T

ask

HR

-C ‘D

efin

ition

of P

re-in

itiat

or H

uman

Fai

lure

Eve

nts’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

HR

-C

Def

initi

on o

f a h

uman

failu

re e

vent

to m

odel

the

impa

ct o

f the

failu

re in

the

syst

em o

r fu

nctio

nal f

ailu

re m

odel

is p

erfo

rmed

with

in th

is ta

sk.

HR

-C01

Fo

r eac

h un

scre

ened

act

ivity

, a h

uman

failu

re e

vent

is d

efin

ed th

at re

pres

ents

the

impa

ct o

f th

e hu

man

failu

re a

t the

leve

l app

ropr

iate

to th

e m

odel

ling

of th

e fu

nctio

n, sy

stem

, or

com

pone

nt(s

) aff

ecte

d.

RA

TIO

NA

LE:

All

sign

ifica

nt c

ontri

buto

rs to

equ

ipm

ent u

nava

ilabi

lity

shou

ld b

e in

clud

ed in

the

syst

em m

odel

s.

EXA

MPL

E: if

the

cons

eque

nce

of th

e hu

man

failu

re is

to m

ake

a tra

in o

f a

syst

em u

nava

ilabl

e, th

en th

e H

FE w

ill b

e in

clud

ed in

the

syst

em m

odel

as

a b

asic

eve

nt re

pres

entin

g th

e fa

ilure

of t

hat t

rain

from

the

hum

an

caus

e. I

f a m

isca

libra

tion

of to

rque

switc

hes p

oten

tially

aff

ects

a n

umbe

r of

val

ves,

then

the

sam

e H

FE w

ould

be

incl

uded

as a

cau

se o

f fai

lure

of

each

of t

hose

val

ves w

here

ver t

hey

appe

ar in

the

syst

em m

odel

s.

70

T

able

8.2

-D

Att

ribu

tes f

or H

uman

Rel

iabi

lity

Ana

lysi

s: T

ask

HR

-D ‘A

sses

smen

t of P

roba

bilit

ies o

f Pre

-initi

ator

Hum

an F

ailu

re E

vent

s’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

HR

-D

Ass

essm

ent o

f the

pro

babi

litie

s of t

he p

re-in

itiat

or h

uman

failu

re e

vent

s is p

erfo

rmed

in a

co

nsis

tent

way

aim

ed to

ass

ure

the

reas

onab

lene

ss o

f fin

al h

uman

err

or p

roba

bilit

ies

(HEP

s).

HR

-D01

Th

e pr

obab

ilitie

s of t

he h

uman

failu

re e

vent

s are

est

imat

ed u

sing

a sy

stem

atic

pro

cess

, so

that

the

estim

atio

n is

per

form

ed o

n a

cons

iste

nt b

asis

. Th

is is

impo

rtant

to e

nabl

e th

e H

FEs t

o be

incl

uded

so th

at th

eir r

elat

ive

impo

rtanc

e is

pre

serv

ed.

CO

MM

ENT:

An

acce

ptab

le m

odel

for e

stim

atin

g H

EPs i

s TH

ERP

(see

R

ef. [

14])

.

HR

-D02

Sc

reen

ing

valu

es o

f the

pro

babi

litie

s of t

he H

FEs,

base

d on

a si

mpl

e m

odel

of t

he g

ener

al

proc

edur

es fo

r res

tora

tion,

are

use

d. W

hen

cons

truct

ing

mod

els f

or e

stim

atin

g th

e H

EPs,

the

char

acte

ristic

s of v

erifi

catio

n pr

oces

ses a

re c

onsi

dere

d, in

clud

ing:

−

Use

of a

writ

ten

chec

k-of

f lis

t; −

Inde

pend

ence

of v

erifi

catio

n of

stat

us;

− Pe

rfor

man

ce o

f a fu

ll fu

nctio

nal t

est b

efor

e th

e ac

tivity

is c

onsi

dere

d co

mpl

ete;

−

Freq

uenc

y of

ver

ifica

tion

of st

atus

com

pare

d to

freq

uenc

y of

per

form

ance

of t

he

activ

ity.

RA

TIO

NA

LE:

Typi

cally

, una

vaila

bilit

y du

e to

failu

re to

reco

nfig

ure

is

not a

maj

or c

ontri

buto

r whe

n co

mpa

red

with

oth

er m

odes

of

unav

aila

bilit

y. T

here

fore

, in

the

case

that

the

HFE

s are

incl

uded

in th

e m

odel

, e.g

. for

com

plet

enes

s, a

deta

iled

mod

el is

unn

eces

sary

for m

any

appl

icat

ions

, and

a si

mpl

e sc

reen

ing

estim

atio

n w

ill su

ffic

e.

HR

-D03

A

det

aile

d as

sess

men

t of t

he H

EP is

per

form

ed fo

r eac

h H

FE fo

r whi

ch th

e sc

reen

ing

HEP

is c

ompa

rabl

e to

the

prob

abili

ties o

f oth

er m

odes

of u

nava

ilabi

lity.

A d

etai

led

asse

ssm

ent r

equi

res a

mor

e th

orou

gh in

vest

igat

ion

of th

e pl

ant p

ract

ices

and

con

ditio

ns

(e.g

. det

ails

of w

ritte

n pr

oced

ures

and

pla

nt la

yout

) and

resu

lts in

a h

ighe

r con

fiden

ce in

th

e de

term

inat

ion

of th

e si

gnifi

canc

e to

risk

of t

he p

re-in

itiat

ing

even

t act

iviti

es. T

he

deta

iled

asse

ssm

ent c

onsi

ders

the

spec

ific

aspe

cts o

f the

pro

cedu

res a

nd th

e m

an-m

achi

ne

inte

rfac

e.

RA

TIO

NA

LE:

Whe

n th

e un

avai

labi

lity

to re

conf

igur

e ev

alua

ted

usin

g a

sim

ple

mod

el is

com

para

ble

with

the

prob

abili

ties o

f oth

er m

odes

of

unav

aila

bilit

y, a

det

aile

d as

sess

men

t sho

uld

be p

erfo

rmed

.

HR

-D04

Th

e de

gree

of d

epen

denc

e be

twee

n H

FEs i

s ass

esse

d ta

king

into

con

side

ratio

n w

heth

er

ther

e ar

e co

mm

on e

lem

ents

in th

eir c

ause

.

CO

MM

ENT:

Som

e an

alys

ts w

ill a

rgue

that

such

dep

ende

nt e

ffect

s are

in

clud

ed in

the

com

mon

cau

se fa

ilure

eve

nts c

onsi

dere

d fo

r the

aff

ecte

d co

mpo

nent

s. It

is im

porta

nt, w

hen

such

a c

laim

is m

ade,

to e

nsur

e th

at

this

is in

deed

the

case

. In

any

cas

e, th

e qu

alita

tive

part

of th

e as

sess

men

t su

gges

ted

here

pro

vide

s pot

entia

lly v

alua

ble

insi

ghts

.

EXA

MPL

ES o

f com

mon

ele

men

ts: t

he p

erfo

rman

ce o

f the

sam

e ac

tivity

on

diff

eren

t tra

ins o

f the

sam

e sy

stem

by

the

sam

e m

aint

enan

ce p

erso

nnel

in

the

sam

e tim

e fr

ame;

the

use

of a

com

mon

pro

cedu

re

71

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

HR

-D05

T

he u

ncer

tain

ties i

n th

e H

EP e

stim

ates

are

cha

ract

eriz

ed c

onsi

sten

t with

the

data

base

us

ed.

RA

TIO

NA

LE: B

ecau

se th

ere

is li

ttle

actu

aria

l dat

a on

HEP

s, th

ey w

ill b

e un

certa

in.

An

asse

ssm

ent o

f the

unc

erta

inty

is im

porta

nt w

hen

usin

g th

e re

sults

of a

PSA

so th

at th

e se

nsiti

vity

of a

ny c

oncl

usio

ns d

raw

n fr

om th

e PS

A c

an b

e ex

amin

ed.

HR

-D06

Th

e re

ason

able

ness

of t

he e

stim

ates

is c

heck

ed b

y co

nfirm

ing

that

ther

e ha

s not

bee

n a

sign

ifica

nt h

isto

ry o

f res

tora

tion

failu

res o

r mis

calib

ratio

n is

sues

. R

ATI

ON

ALE

: Ev

iden

ce o

f a h

isto

ry o

f pro

blem

s poi

nts t

o th

e ne

ed fo

r a

mor

e de

taile

d ex

amin

atio

n. L

ack

of su

ch a

his

tory

pro

vide

s sup

port

for

the

mod

ellin

g as

sum

ptio

ns.

Sp

ecia

l Attr

ibut

e H

R-D

06-S

1:

The

data

col

lect

ed fo

r the

par

amet

er e

stim

atio

n ta

sk is

revi

ewed

for

evid

ence

of o

ccur

renc

es o

f mis

alig

nmen

t or m

isca

libra

tion

prob

lem

s.

The

resu

lts o

f the

revi

ew a

re u

sed

to c

onfir

m th

e re

ason

able

ness

of t

he

estim

ates

.

RATI

ON

ALE:

A m

ore

com

plet

e in

vest

igat

ion

of th

e pl

ant h

isto

ry c

an b

e ob

tain

ed b

y lo

okin

g th

roug

h th

e pl

ant m

aint

enan

ce re

cord

s.

72

T

able

8.2

-E

Att

ribu

tes f

or H

uman

Rel

iabi

lity

Ana

lysi

s: T

ask

HR

-E ‘I

dent

ifica

tion

of P

ost-

initi

ator

Ope

rato

r R

espo

nses

’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

HR

-E

Iden

tific

atio

n of

the

set o

f ope

rato

r res

pons

es fo

llow

ing

the

initi

atin

g ev

ent r

equi

red

for

each

of t

he a

ccid

ent s

eque

nces

mod

elle

d is

per

form

ed in

a sy

stem

atic

way

usi

ng

appr

opria

te in

form

atio

n so

urce

s.

HR

-E01

Th

e se

t of o

pera

tor r

espo

nses

requ

ired

to c

ontro

l and

safe

ly sh

utdo

wn

the

plan

t fol

low

ing

an in

itiat

ing

even

t is g

ener

ated

by

revi

ewin

g al

l rel

evan

t ope

ratin

g pr

oced

ures

(e.g

. em

erge

ncy

oper

atin

g pr

oced

ures

, abn

orm

al o

pera

ting

proc

edur

es, a

nnun

ciat

or re

spon

se

proc

edur

es) t

o de

term

ine

wha

t act

ions

are

requ

ired

as a

func

tion

of th

e pl

ant s

tatu

s re

pres

ente

d in

the

deve

lopm

ent o

f the

acc

iden

t seq

uenc

es. (

See

also

Tab

le 5

.2-C

, Gen

eral

A

ttrib

ute

AS-

C05

).

The

follo

win

g op

erat

or re

spon

ses a

re id

entif

ied:

-

Act

ions

requ

ired

to in

itiat

e, c

ontro

l, is

olat

e, o

r ter

min

ate

syst

ems a

s req

uire

d to

pre

vent

or

miti

gate

cor

e da

mag

e).

- A

ctio

ns re

quire

d to

cha

nge

the

stat

us o

f com

pone

nts i

n or

der t

o fu

lfil a

func

tion

requ

ired

to p

reve

nt o

r miti

gate

cor

e da

mag

e.

CO

MM

ENT:

Thi

s tas

k is

an

inte

gral

par

t of t

he d

evel

opm

ent o

f the

ac

cide

nt se

quen

ce m

odel

.

EXA

MPL

ES:

- Is

olat

ion

of a

faul

ted

stea

m g

ener

ator

, ini

tiatio

n of

the

RH

R s

yste

m,

depr

essu

rizat

ion

of th

e re

acto

r coo

lant

syst

em (B

WR

).

- O

peni

ng a

PO

RV

blo

ck v

alve

.

Sp

ecia

l Attr

ibut

e H

R-E0

1-S1

:

Actio

ns ta

ken

in th

e co

ntro

l roo

m (o

r oth

er c

ontin

uous

ly m

anne

d st

atio

ns) e

ither

in re

spon

se to

pro

cedu

ral d

irec

tion,

or a

s ski

ll-of

-the-

craf

t, to

reco

ver f

rom

a fa

ilure

of e

quip

men

t to

auto

mat

ical

ly in

itiat

e or

ch

ange

stat

e as

requ

ired

are

iden

tifie

d.

RATI

ON

ALE:

Fai

lure

s of t

he in

itiat

ion

sign

als a

re ty

pica

lly a

smal

l fr

actio

n of

the

com

pone

nt fa

ilure

pro

babi

lity,

so th

at th

ese

reco

very

ac

tions

are

typi

cally

not

sign

ifica

nt c

ontr

ibut

ors t

o C

DF.

How

ever

, the

ir

incl

usio

n le

ads t

o a

mor

e co

mpl

ete

mod

el.

CO

MM

ENT:

Bec

ause

of c

utse

t spe

cific

dep

ende

nce

of a

vaila

ble

time,

th

ese

may

be

best

inco

rpor

ated

in th

e m

odel

solu

tion

as re

cove

ry a

ctio

ns

(see

Tab

le 8

.2-H

, Tas

k H

R-H

).

EXAM

PLE:

Man

ual s

tart

s of a

stan

dby

pum

p fo

llow

ing

failu

re o

f aut

o-st

art.

73

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

Sp

ecia

l Attr

ibut

e H

R-E0

1-S2

: Si

gnifi

cant

err

ors o

f com

mis

sion

, i.e

. act

ions

that

lead

to a

dditi

onal

fu

nctio

nal u

nava

ilabi

litie

s, or

inap

prop

riat

ely

initi

ate

syst

em a

re

iden

tifie

d.

RATI

ON

ALE:

Err

ors o

f com

mis

sion

can

lead

to th

e cr

eatio

n of

add

ition

al

acci

dent

sequ

ence

s.

CO

MM

ENT:

Whi

le it

is n

ot y

et g

ener

al p

ract

ice

to in

clud

e er

rors

of

com

mis

sion

in th

e ba

se P

SA, i

t is a

dvan

tage

ous t

o us

e in

form

atio

n on

the

gene

ral c

ause

s of e

rror

s of c

omm

issi

on (s

ee fo

r exa

mpl

e, N

URE

G-1

624,

Re

f. [1

5]) t

o re

duce

the

pote

ntia

l for

intr

oduc

ing

chan

ges t

hat c

ould

in

crea

se th

e lik

elih

ood

of, o

r cre

ate

cond

ition

s con

duci

ve to

, err

ors o

f co

mm

issi

on. T

he m

etho

dolo

gy fo

r the

iden

tific

atio

n of

err

ors o

f co

mm

issi

on is

not

as w

ell f

orm

ulat

ed a

s tha

t of e

rror

s of o

mis

sion

. An

exam

ple

of a

n ap

proa

ch is

ATH

EAN

A (R

ef. [

15])

.

EXAM

PLE:

An

erro

r of c

omm

issi

on m

ight

be

secu

ring

an

inje

ctio

n sy

stem

inap

prop

riat

ely.

HR

-E02

Th

e pr

oced

ures

are

revi

ewed

with

pla

nt o

pera

tions

and

trai

ning

per

sonn

el to

con

firm

that

th

e in

terp

reta

tion

of th

e pr

oced

ures

, and

the

expe

cted

resp

onse

s are

con

sist

ent w

ith tr

aini

ng

and

plan

t ope

ratio

nal p

ract

ices

. Fo

r the

mor

e ch

alle

ngin

g se

quen

ces,

a de

taile

d ta

lk-

thro

ugh

of th

e de

velo

pmen

t of t

he se

quen

ces i

s per

form

ed (s

ee a

lso

Tabl

e 5.

2-A

, Gen

eral

A

ttrib

ute

AS-

C04

).

RA

TIO

NA

LE:

The

writ

ten

proc

edur

es m

ay n

ot a

lway

s giv

e a

pref

eren

ce

of th

e or

der i

n w

hich

the

vario

us o

ptio

ns a

vaila

ble

to p

rovi

de fo

r a

requ

ired

safe

ty fu

nctio

n ar

e ex

erci

sed.

HR

-E03

Si

mul

ator

exe

rcis

es a

re o

bser

ved

to g

ain

a ge

nera

l und

erst

andi

ng o

f cre

w d

ynam

ics,

and

the

use

of th

e pr

oced

ures

.

For t

he m

ore

chal

leng

ing

acci

dent

sequ

ence

s, si

mul

ator

exe

rcis

es a

re o

bser

ved.

Suc

h ex

erci

ses g

ive

info

rmat

ion

on p

erfo

rman

ce sh

apin

g fa

ctor

s, su

ch a

s the

pre

senc

e of

di

stra

ctin

g an

nunc

iato

rs, t

he ti

min

g as

soci

ated

with

the

actio

ns, t

he c

ompl

exity

of t

he

actio

ns, e

tc.

RA

TIO

NA

LE:

An

unde

rsta

ndin

g of

the

perf

orm

ance

shap

ing

fact

ors i

s im

porta

nt fo

r the

eva

luat

ion

of th

e H

EPs (

see

Tabl

e 8.

2-F,

Tas

k H

R-F

, and

Ta

ble

8.2-

G, T

ask

HR

-G).

74

T

able

8.2

-F

Att

ribu

tes f

or H

uman

Rel

iabi

lity

Ana

lysi

s: T

ask

HR

-F ‘D

efin

ition

of P

ost-

initi

ator

Hum

an F

ailu

re E

vent

s’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

HR

-F

The

task

con

sist

s of t

he d

efin

ition

of h

uman

failu

re e

vent

s tha

t rep

rese

nt th

e fa

ilure

to

resp

ond

as re

quire

d th

at a

re c

onsi

sten

t with

the

stru

ctur

e an

d le

vel o

f det

ail i

n th

e ac

cide

nt

sequ

ence

mod

els.

HR

-F01

H

uman

failu

re e

vent

s rep

rese

ntin

g th

e im

pact

of f

ailu

re to

per

form

a re

quire

d fu

nctio

n ar

e id

entif

ied

and

incl

uded

in th

e pl

ant l

ogic

mod

el.

Failu

res t

o pe

rfor

m m

ore

than

one

re

spon

se a

re g

roup

ed in

to a

sing

le H

FE, i

f the

impa

ct o

n th

e ac

cide

nt se

quen

ce

deve

lopm

ent i

s the

sam

e or

can

be

boun

ded.

EXA

MPL

ES:

An

HFE

may

be

incl

uded

in th

e pl

ant l

ogic

mod

el in

a n

umbe

r of w

ays:

- as

an

even

t tre

e br

anch

poi

nt;

- as

a c

ontri

buto

r to

a fu

nctio

nal l

evel

faul

t tre

e us

ed to

eva

luat

e th

e fa

ilure

of a

func

tion

or sy

stem

;

- in

a sy

stem

faul

t tre

e as

a m

ode

of u

nava

ilabi

lity

of a

com

pone

nt,

segm

ent,

or tr

ain.

HR

-F02

Th

e de

finiti

on o

f eac

h H

FE in

pre

para

tion

for e

stim

atio

n of

its p

roba

bilit

y is

com

plet

ed b

y sp

ecify

ing

the

scen

ario

spec

ific

fact

ors i

nclu

ding

: -

avai

labi

lity

of c

ues a

nd/o

r oth

er in

dica

tions

for a

lerti

ng th

e op

erat

ors t

o th

e ne

ed fo

r ac

tion;

-

the

scen

ario

spec

ific

proc

edur

al g

uida

nce;

-

time

avai

labl

e fo

r suc

cess

ful c

ompl

etio

n of

resp

onse

; -

timin

g of

cue

s rel

ativ

e to

acc

iden

t pro

gres

sion

; -

avai

labi

lity

of sy

stem

s, or

com

pone

nts i

dent

ified

in th

e pr

oced

ures

; -

the

task

s com

pris

ing

the

requ

ired

resp

onse

.

CO

MM

ENTS

: M

any

exis

ting

HR

A m

odel

s (i.e

., m

etho

ds fo

r cal

cula

ting

hum

an e

rror

pro

babi

litie

s) d

o no

t add

ress

som

e of

thes

e fa

ctor

s exp

licitl

y.

How

ever

, it i

s nec

essa

ry to

und

erst

and

the

fact

ors,

at le

ast q

ualit

ativ

ely

so

that

Gen

eral

Attr

ibut

e H

R-G

06 (T

able

8.2

-G) c

an b

e sa

tisfie

d.

75

T

able

8.2

-G

Att

ribu

tes f

or H

uman

Rel

iabi

lity

Ana

lysi

s: T

ask

HR

-G ‘A

sses

smen

t of P

roba

bilit

ies o

f Pos

t-in

itiat

or H

uman

Fai

lure

Eve

nts’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

HR

-G

Ass

essm

ent o

f the

pro

babi

litie

s of t

he p

ost-i

nitia

tor h

uman

failu

re e

vent

s, an

d a

char

acte

rizat

ion

of th

e as

soci

ated

unc

erta

intie

s and

dep

ende

nce

betw

een

the

even

ts a

re

perf

orm

ed u

sing

app

ropr

iate

met

hodo

logi

es a

nd d

ata.

HR

-G01

D

etai

led

asse

ssm

ent o

f the

HEP

s is p

erfo

rmed

for t

he si

gnifi

cant

HFE

s (se

e N

ote)

. Sc

reen

ing

valu

es a

re u

sed

for t

he n

on-s

igni

fican

t HFE

s. C

OM

MEN

TS:

1) T

he id

entif

icat

ion

of th

e si

gnifi

cant

HFE

s is a

n ite

rativ

e pr

oces

s re

quiri

ng se

vera

l qua

ntifi

catio

ns o

f the

PSA

mod

el.

2) A

sign

ifica

nt H

FE is

one

that

con

tribu

tes m

easu

rabl

y to

the

curr

ent

leve

l of r

isk,

or i

s rel

ied

upon

to a

chie

ve th

e cu

rren

t lev

el o

f ris

k. A

n ex

ampl

e of

crit

eria

that

can

be

used

to d

eter

min

e si

gnifi

canc

e is

the

follo

win

g: a

sign

ifica

nt b

asic

eve

nt h

as a

Fus

sell-

Ves

ely

impo

rtanc

e m

easu

re (F

V) g

reat

er th

an a

pre

defin

ed n

umbe

r (i.e

. 0.0

05 o

r 0.0

1), o

r a

Ris

k A

chie

vem

ent W

orth

(RA

W) g

reat

er th

an a

pre

defin

ed n

umbe

r (i.

e. 2

). A

ltern

ativ

e im

porta

nce

mea

sure

s and

crit

eria

may

be

used

.

Sp

ecia

l Attr

ibut

e H

R-G

01-S

1:

A de

taile

d as

sess

men

t is p

erfo

rmed

for e

ach

HFE

. RA

TIO

NAL

E : P

erfo

rmin

g a

deta

iled

asse

ssm

ent o

f all

HFE

s avo

ids t

he

need

for i

tera

tion.

HR

-G02

Th

e m

etho

d us

ed to

ass

ess H

EPs a

ddre

sses

failu

re in

cog

nitio

n (d

etec

tion,

situ

atio

n as

sess

men

t, an

d re

spon

se p

lann

ing)

as w

ell a

s fai

lure

s in

exec

utio

n.

RA

TIO

NA

LE:

Failu

res i

n co

gniti

on c

an th

e m

ore

impo

rtant

con

tribu

tors

to

failu

re, a

nd fu

rther

mor

e, c

an b

e a

sign

ifica

nt so

urce

of d

epen

denc

e be

twee

n H

FEs (

see

HR

-G07

).

Sp

ecia

l Attr

ibut

e H

R-G

02-S

1:

The

met

hod

used

to a

sses

s HEP

s is c

apab

le o

f eva

luat

ing

the

impa

ct o

f pr

oced

ure

chan

ges.

RATI

ON

ALE :

Eva

luat

ion

of th

e im

pact

of p

roce

dura

l cha

nges

is e

ssen

tial

for a

pplic

atio

ns a

imed

to o

ptim

ise

EOPs

and

AM

Ps.

HR

-G03

Th

e m

odel

use

d to

ass

ess t

he H

EPs a

ddre

sses

the

follo

win

g pe

rfor

man

ce sh

apin

g fa

ctor

s on

a sc

enar

io a

nd p

lant

-spe

cific

bas

is:

- th

e ty

pe (c

lass

room

or s

imul

ator

) and

freq

uenc

y of

trai

ning

on

the

resp

onse

; -

the

qual

ity o

f the

writ

ten

proc

edur

es a

nd a

dmin

istra

tive

cont

rols

; -

avai

labi

lity

of n

eces

sary

inst

rum

enta

tion;

-

degr

ee o

f cla

rity

of th

e cu

es/in

dica

tions

; -

timin

g is

sues

- tim

e at

whi

ch c

ues a

re re

ceiv

ed, t

ime

requ

ired

to p

erfo

rm th

e re

spon

se,

and

the

time

avai

labl

e to

com

plet

e th

e re

spon

se;

- co

mpl

exity

of t

he ta

sks t

o be

per

form

ed;

- na

ture

of t

he h

uman

-mac

hine

inte

rfac

e.

CO

MM

ENT:

Man

y H

RA

mod

els d

o no

t dea

l with

eac

h of

thes

e fa

ctor

s ex

plic

itly.

How

ever

, the

y sh

ould

be

take

n in

to a

ccou

nt w

hen

com

parin

g th

e re

lativ

e va

lues

of H

EPs (

see

HR

-G6)

.

The

com

plex

ity m

ay b

e as

sess

ed in

par

t on

the

basi

s of a

task

ana

lysi

s.

Task

ana

lyse

s can

be

perf

orm

ed a

t a n

umbe

r of d

iffer

ent l

evel

s of d

etai

l as

requ

ired

by th

e m

odel

use

d to

qua

ntify

the

HEP

s.

76

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

For a

spec

ific

type

of o

pera

tor r

espo

nse,

thes

e fa

ctor

s can

var

y de

pend

ing

on th

e sc

enar

io

bein

g m

odel

led,

e.g

. a lo

ss o

f a d

c bu

s may

aff

ect t

he a

vaila

bilit

y of

inst

rum

enta

tion.

Whe

n th

e re

spon

se re

quire

s act

ions

out

side

the

cont

rol r

oom

, the

follo

win

g fa

ctor

s are

ta

ken

into

acc

ount

in a

dditi

on to

the

abov

e:

- en

viro

nmen

tal f

acto

rs (e

.g. h

eat,

radi

atio

n, h

umid

ity);

- ac

cess

ibili

ty o

f equ

ipm

ent t

o be

man

ipul

ated

; -

need

for s

peci

al to

ols;

-

time

to re

ach

phys

ical

ly th

e pl

ace

if no

t per

man

ently

occ

upie

d;

- co

mm

unic

atio

n is

sues

bet

wee

n M

CR

and

loca

l per

sonn

el.

HR

-G04

Th

e tim

elin

e fo

r occ

urre

nce

of c

ues a

nd th

e ev

alua

tion

of th

e tim

e av

aila

ble

to c

ompl

ete

the

actio

n is

bas

ed o

n ap

plic

able

gen

eric

stud

ies (

e.g.

ther

mal

-hyd

raul

ic a

naly

ses)

. R

ATI

ON

ALE

: D

epen

ding

on

the

met

hod

used

to e

stim

ate

the

HEP

s, a

prec

ise

eval

uatio

n of

the

times

may

not

be

nece

ssar

y. F

or th

ose

that

are

ba

sed

prim

arily

on

time,

the

Spec

ial A

ttrib

ute

appl

ies.

Sp

ecia

l Attr

ibut

e H

R-G

04-S

1:

The

time

line

is b

ased

on

plan

t-spe

cific

ther

mal

-hyd

raul

ic a

naly

ses.

HR

-G05

Fo

r tho

se H

FEs f

or w

hich

a d

etai

led

eval

uatio

n is

per

form

ed, t

he ti

me

to c

ompl

ete

the

actio

n is

bas

ed o

n ac

tual

tim

e m

easu

rem

ents

in w

alkt

hrou

ghs,

talk

-thro

ughs

of t

he

proc

edur

es, o

r sim

ulat

or o

bser

vatio

ns.

RA

TIO

NA

LE:

Dep

endi

ng o

n th

e m

etho

d us

ed to

est

imat

e th

e H

EPs,

a pr

ecis

e ev

alua

tion

of th

e tim

es m

ay n

ot b

e ne

cess

ary.

How

ever

, as a

m

inim

um, a

n es

timat

e of

the

time

is n

eede

d to

con

firm

the

feas

ibili

ty o

f th

e ac

tions

.

HR

-G06

A

fter e

stim

atio

n of

the

HEP

s, th

e re

lativ

e va

lues

are

ass

esse

d to

che

ck th

at th

ey a

re

refle

ctiv

e of

the

impa

ct o

f the

var

ious

per

form

ance

-sha

ping

fact

ors i

dent

ified

in H

R-G

03

abov

e.

RA

TIO

NA

LE:

This

is a

cru

cial

attr

ibut

e. B

ecau

se o

f the

lack

of a

ctua

rial

data

, HEP

s will

alw

ays b

e un

certa

in.

This

mus

t be

take

n in

to a

ccou

nt

whe

n in

terp

retin

g th

e re

sults

of t

he P

SA.

How

ever

, it i

s ess

entia

l tha

t the

re

lativ

e va

lues

of t

he H

EPs b

e re

cogn

ized

as t

hey

can

affe

ct th

e re

lativ

e ris

k si

gnifi

canc

e of

acc

iden

t seq

uenc

es, f

unct

ions

, sys

tem

s, an

d co

mpo

nent

s, w

hich

are

impo

rtant

for m

any

appl

icat

ions

.

77

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

HR

-G07

Th

e de

gree

of d

epen

denc

e be

twee

n H

FEs a

ppea

ring

in th

e sa

me

acci

dent

sequ

ence

or c

ut

set i

s ass

esse

d. F

acto

rs a

ffect

ing

the

degr

ee o

f dep

ende

nce

incl

ude:

-

use

of c

omm

on c

ues;

-

resp

onse

s cal

led

for i

n th

e sa

me

proc

edur

e;

- cl

osen

ess i

n tim

e of

cue

s or r

equi

red

actio

ns;

- in

crea

sed

stre

ss c

ause

d by

failu

re o

f the

firs

t res

pons

e.

A c

ondi

tiona

l pro

babi

lity

of th

e se

cond

, thi

rd, e

tc. e

vent

, giv

en fa

ilure

of t

he fi

rst,

seco

nd,

etc.

is e

valu

ated

. Th

e as

sum

ptio

n of

inde

pend

ence

bet

wee

n H

FEs i

s jus

tifie

d.

RA

TIO

NA

LE:

This

is a

cru

cial

attr

ibut

e. B

ecau

se a

ccid

ent s

eque

nce

mod

els a

re d

evel

oped

in te

rms o

f dis

cret

e fu

nctio

ns o

r in

term

s of s

epar

ate

syst

ems,

cons

ider

atio

n of

HFE

s for

eac

h fu

nctio

n or

syst

em in

isol

atio

n ca

n le

ad to

exc

essi

ve c

redi

t for

ope

rato

r act

ion

unle

ss th

e de

pend

ency

is

acco

unte

d fo

r.

CO

MM

ENT:

The

ass

umpt

ion

of in

depe

nden

ce is

supp

orte

d by

arg

umen

ts

such

as:

-

the

requ

ired

actio

ns a

re se

para

ted

suffi

cien

tly in

the

deve

lopm

ent o

f th

e ac

cide

nt se

quen

ce;

- th

e cu

es fo

r sub

sequ

ent a

ctio

ns a

re in

depe

nden

t of t

hose

for p

rior

actio

ns;

- th

e w

orkl

oad

is n

ot si

gnifi

cant

ly in

crea

sed

by v

irtue

of t

he fa

ilure

of

the

prio

r act

ions

; -

the

seco

nd a

ctio

n w

ould

be

requ

ired

whe

ther

or n

ot th

e fir

st w

ere

requ

ired.

HR

-G08

A

cha

ract

eriz

atio

n of

the

unce

rtain

ty in

the

HEP

s is p

rovi

ded,

con

sist

ent w

ith th

e qu

antif

icat

ion

met

hod.

R

ATI

ON

ALE

: Bec

ause

ther

e is

littl

e ac

tuar

ial d

ata

on H

EPs,

they

will

be

unce

rtain

. A

n as

sess

men

t of t

he u

ncer

tain

ty is

impo

rtant

whe

n us

ing

the

resu

lts o

f a P

SA so

that

the

robu

stne

ss o

f any

con

clus

ions

dra

wn

from

the

PSA

can

be

exam

ined

.

78

T

able

8.2

-H

Att

ribu

tes f

or H

uman

Rel

iabi

lity

Ana

lysi

s: T

ask

HR

-H ‘R

ecov

ery

Act

ions

’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

HR

-H

The

task

incl

udes

iden

tific

atio

n of

add

ition

al re

cove

ry a

ctio

ns a

fter t

he in

itial

solu

tion

of

the

PSA

mod

el.

HR

-H01

R

ecov

ery

actio

ns th

at c

an re

stor

e th

e fu

nctio

ns o

f sys

tem

s or c

ompo

nent

s are

incl

uded

on

an a

s-ne

eded

bas

is to

elim

inat

e un

nece

ssar

ily c

onse

rvat

ive

cont

ribut

ions

to a

ccid

ent

sequ

ence

s.

CO

MM

ENTS

:

1) U

se o

f div

erse

or a

ltern

ativ

e m

eans

is c

onsi

dere

d as

reco

very

act

ions

.

2) T

he re

cove

ry a

ctio

ns a

re in

clud

ed in

the

mod

el a

t a le

vel (

e.g.

scen

ario

, cu

tset

) suc

h th

at th

e co

ntex

t (e.

g. P

SFs)

ass

ocia

ted

with

that

leve

l, w

hich

det

erm

ines

the

prob

abili

ty o

f the

reco

very

act

ion,

can

be

rega

rded

as u

nifo

rm.

For e

xam

ple,

such

act

ions

are

ofte

n in

clud

ed a

t th

e cu

tset

leve

l, be

caus

e th

e co

ntex

t (e.

g. ti

me

avai

labl

e to

per

form

the

actio

n) c

an v

ary

from

cut

set t

o cu

tset

.

HR

-H02

R

ecov

ery

actio

ns a

re c

redi

ted

only

if:

− a

proc

edur

e is

ava

ilabl

e an

d op

erat

or tr

aini

ng h

as b

een

prov

ided

for t

he a

ctio

n O

R it

is

cons

ider

ed to

be

a sk

ill-o

f-the

-cra

ft ac

tion

, i.e

., on

e fo

r whi

ch a

pro

cedu

re is

not

nee

ded

as it

is o

ne th

at, b

ecau

se o

f the

ope

rato

r’s s

kill

and

expe

rienc

e, is

cle

arly

iden

tifia

ble,

A

ND

−

cues

(e.g

. an

annu

ncia

tor)

ale

rts th

e op

erat

ors t

o th

e ne

ed fo

r act

ion

OR

the

proc

edur

e di

rect

s the

ope

rato

r to

chec

k th

e st

atus

of t

he c

ompo

nent

,

AN

D

− fe

asib

ility

of t

he a

ctio

n is

con

firm

ed.

HR

-H03

Th

e po

tent

ial f

or d

epen

denc

y be

twee

n re

cove

ry a

ctio

ns a

nd a

ny o

ther

HFE

s in

the

acci

dent

se

quen

ce c

utse

t is a

sses

sed.

R

ATI

ON

ALE

: Th

e ne

ed fo

r rec

over

y ac

tions

is g

ener

ally

reco

gniz

ed b

y th

e co

ntro

l roo

m c

rew

resp

onsi

ble

for t

he p

erfo

rman

ce o

f the

pr

oced

ural

ized

act

ions

mod

elle

d in

the

even

t tre

es a

nd fa

ult t

rees

. Th

eref

ore,

ther

e m

ay b

e pe

rfor

man

ce sh

apin

g fa

ctor

s tha

t aff

ect b

oth

reco

very

and

pro

cedu

raliz

ed a

ctio

ns.

79

T

able

8.2

-I

Att

ribu

tes f

or H

uman

Rel

iabi

lity

Ana

lysi

s: T

ask

HR

-I ‘

Doc

umen

tatio

n’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

HR

-I

Doc

umen

tatio

n an

d in

form

atio

n st

orag

e is

ade

quat

ely

perf

orm

ed.

HR

-I01

Th

e H

RA

ana

lysi

s is d

ocum

ente

d in

eno

ugh

deta

il to

repr

oduc

e re

sults

and

per

mit

the

revi

ewer

s to

unde

rsta

nd li

mita

tions

impo

sed

by th

e m

odel

s, as

sum

ptio

ns, a

nd d

ata.

The

sp

ecifi

c as

pect

s of t

he H

RA

to b

e do

cum

ente

d in

clud

e th

e fo

llow

ing:

− H

RA

met

hodo

logy

: •

App

roac

h us

ed to

iden

tify

HFE

s •

Met

hods

use

d to

est

imat

e H

EPs

• A

ppro

ach

to th

e as

sess

men

t of d

epen

denc

y −

Bas

is fo

r eac

h H

EP:

• Sc

reen

ing

valu

es

• D

etai

led

HEP

eva

luat

ions

: -

Fact

ors u

sed

in th

e qu

antif

icat

ion

of th

e H

EPs

- H

ow th

ey w

ere

char

acte

rized

-

How

they

wer

e in

corp

orat

ed in

the

quan

tific

atio

n

80

9. PSA ELEMENT ‘DA’: DATA ANALYSIS 9.1. Main objectives

The objective of the data analysis is to provide estimates for the parameters, called reliability parameters, of the reliability models specified under systems analysis. The reliability models serve to determine the probabilities of the basic events representing specific equipment failures and unavailabilities. Reliability parameters typically include: - Failure rates

- Probabilities for failure on demand

- Unavailabilities due to maintenance or test, or

- Test and maintenance frequencies and repair, test, and maintenance durations

- Mission times as specified in systems analysis

- Common cause failure (CCF) model parameters.

Important aspects and characteristics of reliability parameters are the following: a) Parameters, whether estimated on the basis of plant-specific or generic data, or both,

appropriately reflect design and operational features of the plant.

b) Component or system unavailabilities due to repair, test and maintenance are properly accounted for.

c) Uncertainties in the data are understood and accounted for.

Note: The parameters of reliability models as discussed here should not be mixed up with the parameters of probability distributions used to describe the uncertainty of reliability parameters.

9.2. Data analysis tasks and their attributes

Table 9.2 lists the main tasks for the PSA element ‘Data Analysis’. Tables 9.2-A

through 9.2-H present the description of general attributes and special attributes for these tasks.

81

Table 9.2 Main Tasks for Data Analysis


DA-A Reliability Model Parameter Identification

DA-B Component Grouping for Parameter Estimation

DA-C Collecting and Evaluating Generic Information

DA-D Plant-Specific Data Collection and Evaluation

DA-E Derivation of Plant-Specific Parameters, Integration of Generic and Plant Specific Information

DA-F Derivation of Plant-Specific Parameters for Common Cause Failure Events

DA-G Handling of Reliability Parameters Affected by Plant Changes, Handling of New Equipment

DA-H Documentation

82

T

able

9.2

-A

Att

ribu

tes f

or D

ata

Ana

lysi

s: T

ask

DA

-A ‘R

elia

bilit

y M

odel

Par

amet

er Id

entif

icat

ion’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

DA

-A

Each

relia

bilit

y pa

ram

eter

is id

entif

ied

for t

he re

liabi

lity

mod

els a

s def

ined

in sy

stem

s an

alys

is in

term

s of t

he lo

gic

failu

re m

odel

(fau

lt tre

es, e

vent

tree

s or s

peci

al fa

ilure

mod

el)

and

basi

c ev

ent c

hara

cter

istic

s and

bou

ndar

y. T

he re

liabi

lity

mod

els a

nd th

eir p

aram

eter

s ar

e id

entif

ied

unde

r Tas

k SY

-B in

Sec

tion

7, T

able

7.2

-B (F

ailu

re C

ause

Iden

tific

atio

n an

d M

odel

ling)

.

DA

-A01

Fr

om sy

stem

s ana

lysi

s, th

e re

liabi

lity

mod

els f

or w

hich

par

amet

ers a

re re

quire

d ar

e id

entif

ied.

Ass

ocia

ted

equi

pmen

t bou

ndar

ies,

failu

re m

odes

, and

mis

sion

succ

ess c

riter

ia a

re

iden

tifie

d co

nsis

tent

with

the

corr

espo

ndin

g ba

sic

even

t def

initi

ons i

n sy

stem

s ana

lysi

s for

fa

ilure

rate

s, fa

ilure

pro

babi

litie

s, un

avai

labi

litie

s, an

d co

mm

on c

ause

failu

re p

aram

eter

s. B

ound

arie

s of u

nava

ilabi

lity

even

ts a

re d

efin

ed c

onsi

sten

t with

cor

resp

ondi

ng d

efin

ition

s in

syst

ems a

naly

sis.

EXA

MPL

E fo

r a c

ompo

nent

mis

sion

succ

ess c

riter

ion:

Min

imal

flow

rate

fo

r 2 h

ours

.

Bas

ic e

vent

s and

ass

ocia

ted

relia

bilit

y m

odel

s for

whi

ch p

aram

eter

s are

re

quire

d ty

pica

lly in

clud

e:

(a) I

ndep

ende

nt o

r com

mon

cau

se fa

ilure

of a

com

pone

nt o

r sys

tem

to

star

t or c

hang

e st

ate

on d

eman

d,

(b) I

ndep

ende

nt o

r com

mon

cau

se fa

ilure

of a

com

pone

nt o

r sys

tem

to

cont

inue

ope

ratin

g or

pro

vide

a re

quire

d fu

nctio

n fo

r a d

efin

ed ti

me

perio

d,

(c) E

quip

men

t una

vaila

bilit

y to

per

form

its r

equi

red

func

tion

due

to b

eing

ou

t of s

ervi

ce fo

r rep

air o

r mai

nten

ance

,

(d) E

quip

men

t una

vaila

bilit

y to

per

form

its r

equi

red

func

tion

due

to b

eing

te

sted

,

(e) F

ailu

re to

reco

ver a

func

tion

or sy

stem

(e.g

. fai

lure

to re

cove

r off

site

-po

wer

),

(f) F

ailu

re to

repa

ir a

com

pone

nt, s

yste

m, o

r fun

ctio

n in

a d

efin

ed ti

me

perio

d.

DA

-A02

B

asic

eve

nt ID

s are

est

ablis

hed

base

d on

the

plan

ts sy

stem

and

equ

ipm

ent I

D sy

stem

as f

ar

as fe

asib

le. T

hus,

equi

pmen

t and

syst

em ID

s are

nor

mal

ly c

onta

ined

in a

ssoc

iate

d ba

sic

even

t ID

s. D

evia

tions

from

this

prin

cipl

e ar

e ju

stifi

ed a

nd th

e ex

act c

orre

latio

n be

twee

n ba

sic

even

ts a

nd p

lant

equ

ipm

ent i

s est

ablis

hed

and

docu

men

ted.

A d

atab

ase

of b

asic

eve

nts a

nd a

ssoc

iate

d re

liabi

lity

mod

els i

s est

ablis

hed

cont

aini

ng th

e ex

act c

orre

latio

n to

spec

ific

plan

t equ

ipm

ent.

83

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

DA

-A03

Th

e pa

ram

eter

s to

be e

stim

ated

and

the

data

requ

ired

are

iden

tifie

d.

The

para

met

ers i

dent

ified

and

thei

r cha

ract

eris

tics a

re in

clud

ed in

the

data

base

est

ablis

hed

unde

r DA

-A02

.

EXA

MPL

ES a

re a

s fol

low

s:

(a) F

or fa

ilure

s on

dem

and

the

para

met

er is

the

prob

abili

ty o

f fai

lure

or

unav

aila

bilit

y on

dem

and,

and

the

data

requ

ired

are

the

num

ber o

f fa

ilure

s giv

en a

num

ber o

f dem

ands

. Bas

ical

ly th

is re

pres

enta

tion

corr

espo

nds t

o a

bino

mia

l pro

babi

lity

mod

el.

(b) F

or st

andb

y fa

ilure

s (if

stan

dby

failu

res a

re d

escr

ibed

in th

is m

anne

r)

and

oper

atin

g fa

ilure

s, th

e pa

ram

eter

is th

e fa

ilure

rate

, and

the

data

re

quire

d ar

e th

e nu

mbe

r of f

ailu

res i

n th

e to

tal (

stan

dby

or o

pera

ting)

tim

e. B

asic

ally

, thi

s rep

rese

ntat

ion

corr

espo

nds t

o a

Pois

son

prob

abili

ty m

odel

.

84

T

able

9.2

-B

Att

ribu

tes f

or D

ata

Ana

lysi

s: T

ask

DA

-B ‘C

ompo

nent

Gro

upin

g fo

r Pa

ram

eter

Est

imat

ion’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

DA

-B

Com

pone

nts a

re g

roup

ed in

to a

ppro

pria

te p

opul

atio

n gr

oups

for p

aram

eter

est

imat

ion.

DA

-B01

Th

e ra

tiona

le fo

r gro

upin

g co

mpo

nent

s int

o a

hom

ogen

eous

pop

ulat

ion

for p

aram

eter

es

timat

ion

cons

ider

s the

des

ign,

env

ironm

enta

l, fu

nctio

nal a

nd o

pera

tiona

l con

ditio

ns o

f the

co

mpo

nent

s in

the

as-b

uilt

and

as-o

pera

ted

plan

t.

For p

aram

eter

est

imat

ion,

com

pone

nts a

re g

roup

ed a

ccor

ding

to ty

pe (e

.g. m

otor

ope

rate

d pu

mp,

air-

oper

ated

val

ve) a

nd a

ccor

ding

to th

e de

taile

d ch

arac

teris

tics o

f the

ir us

age:

a) D

esig

n/si

ze

b) S

yste

m c

hara

cter

istic

s:

- st

andb

y, o

pera

ting

- o

pera

tiona

l con

ditio

ns (e

.g. c

lean

vs.

untre

ated

wat

er, a

ir)

- m

aint

enan

ce p

ract

ices

-

freq

uenc

y of

dem

ands

c)

Env

ironm

enta

l con

ditio

ns

d) O

ther

app

ropr

iate

cha

ract

eris

tics i

nclu

ding

man

ufac

ture

r.

Not

incl

uded

in th

e de

finiti

on o

f a g

roup

are

obv

ious

out

liers

(e.g

. val

ves t

hat a

re n

ever

te

sted

and

unl

ikel

y to

be

oper

ated

are

not

gro

uped

toge

ther

with

thos

e th

at a

re te

sted

or

othe

rwis

e m

anip

ulat

ed fr

eque

ntly

). Th

e gr

oupi

ng a

nd th

e gr

oupi

ng ra

tiona

le a

re in

clud

ed in

th

e da

taba

se c

reat

ed u

nder

Tas

k D

A-A

(Tab

le 9

.2-A

).

EXA

MPL

E: F

unct

iona

l and

ope

ratio

nal c

ondi

tions

rega

rdin

g ce

ntrif

ugal

pu

mps

. Diff

eren

t par

amet

ers c

an b

e de

fined

for l

ow p

ress

ure

pum

ps w

ith

diffe

rent

func

tiona

l and

ope

ratio

nal c

ondi

tions

:

- on

-line

coo

ling

wat

er sy

stem

s whi

ch a

re re

quire

d al

so p

ost t

rip v

ersu

s co

olin

g sy

stem

s nor

mal

ly in

stan

dby;

- co

olin

g w

ater

syst

ems c

ircul

atin

g ra

w w

ater

ver

sus c

oolin

g w

ater

sy

stem

s circ

ulat

ing

clea

n w

ater

;

- w

ell w

ater

pum

ps.

CO

MM

ENT:

A to

o na

rrow

pop

ulat

ion

in a

gro

up m

ay le

ad to

a sa

mpl

e of

pl

ant-s

peci

fic d

ata

that

is n

ot st

atis

tical

ly si

gnifi

cant

.

Sp

ecia

l Attr

ibut

e D

A-B0

1-S1

:

For s

peci

fic a

pplic

atio

ns, s

uch

as th

e as

sess

men

t of t

est p

roce

dure

s for

ce

rtai

n ty

pes o

f pum

ps, a

refin

emen

t of t

he c

ompo

nent

gro

upin

g is

ad

visa

ble.

Thi

s ref

inem

ent p

rovi

des t

he re

solu

tion

requ

ired

for t

hese

sp

ecifi

c ap

plic

atio

ns.

RATI

ON

ALE:

A to

o br

oad

popu

latio

n w

ithin

a g

roup

may

mas

k or

av

erag

e-ou

t the

spec

ific

feat

ures

of p

artic

ular

com

pone

nts.

This

in tu

rn

coul

d si

gnifi

cant

ly h

inde

r cer

tain

app

licat

ions

.

85

T

able

9.2

-C

Att

ribu

tes f

or D

ata

Ana

lysi

s: T

ask

DA

-C ‘C

olle

ctin

g an

d E

valu

atin

g G

ener

ic In

form

atio

n’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

DA

-C

Gen

eric

info

rmat

ion

is c

olle

cted

in a

ccor

danc

e w

ith th

e pa

ram

eter

def

initi

ons o

f Tas

k D

A-

A (9

.2-A

) and

the

grou

ping

ratio

nale

of T

ask

DA

-B (T

able

9.2

-B) a

nd e

valu

ated

rega

rdin

g its

app

licab

ility

. App

ropr

iate

gen

eric

par

amet

ers a

re se

lect

ed o

r par

amet

ers a

re c

ompo

sed

for t

he p

lant

from

app

licab

le g

ener

ic so

urce

s.

DA

-C01

G

ener

ic in

form

atio

n on

failu

res a

nd u

nava

ilabi

litie

s is c

olle

cted

in o

rder

to a

ccou

nt fo

r a

broa

der r

ange

of c

ondi

tions

and

exp

osur

e (e

xpos

ure

time

or to

tal n

umbe

r of d

eman

ds) a

s av

aila

ble

from

the

limite

d pl

ant e

xper

ienc

e. T

he in

form

atio

n m

ay in

clud

e ra

w d

ata,

e.g

. fa

ilure

eve

nts a

nd a

ssoc

iate

d ex

posu

re, o

r may

onl

y be

in th

e fo

rm o

f rel

iabi

lity

mod

el

para

met

ers s

uch

as fa

ilure

rate

s. Th

e so

urce

and

the

deriv

atio

n pr

oces

s of t

he g

ener

ic

para

met

er e

stim

ates

are

iden

tifie

d an

d de

scrib

ed. T

he p

aram

eter

def

initi

ons a

nd b

ound

ary

cond

ition

s are

eva

luat

ed in

vie

w o

f con

sist

ency

with

thos

e de

term

ined

in re

spon

se to

Tas

ks

DA

-A a

nd D

A-B

(Tab

les 9

.2-A

and

9.2

-B re

spec

tivel

y). G

ener

ic d

ata

for u

nava

ilabi

lity

due

to te

st, m

aint

enan

ce, a

nd re

pair

have

to b

e us

ed w

ith c

autio

n si

nce

diffe

rent

pla

nts c

an h

ave

diffe

rent

test

and

mai

nten

ance

phi

loso

phie

s.

Gen

eric

info

rmat

ion

is tr

aced

to th

e pr

imar

y so

urce

to a

void

inad

verte

nt d

oubl

e co

untin

g of

in

form

atio

n an

d to

ass

ure

that

no

info

rmat

ion

from

the

plan

t its

elf i

s con

tain

ed in

the

gene

ric in

form

atio

n in

the

case

that

the

gene

ric a

nd th

e pl

ant s

peci

fic a

re c

ombi

ned

unde

r Ta

sk D

A-E

(Tab

le 9

.2-E

).

The

gene

ric in

form

atio

n is

eva

luat

ed re

gard

ing

its a

pplic

abili

ty fo

r the

pla

nt, c

onsi

derin

g th

e ch

arac

teris

tics,

desi

gn a

nd o

pera

tiona

l fea

ture

s of t

he e

quip

men

t for

whi

ch th

is

info

rmat

ion

is in

tend

ed to

be

appl

ied.

The

colle

ctio

n an

d ev

alua

tion

of g

ener

ic in

form

atio

n in

clud

es a

n un

ders

tand

ing

and

an

asse

ssm

ent o

f the

unc

erta

inty

in th

e or

igin

al d

ata.

EXA

MPL

E: S

ourc

es o

f unc

erta

intie

s reg

ardi

ng th

e us

e of

gen

eric

re

liabi

lity

para

met

er:

- D

iffer

ence

s in

com

pone

nt d

esig

n an

d op

erat

iona

l fea

ture

s

- D

iffer

ence

s in

test

, rep

air a

nd m

aint

enan

ce p

ract

ices

- Q

ualit

y of

the

gene

ric d

ata

(e.g

. com

plet

enes

s)

- St

atis

tical

unc

erta

intie

s

DA

-C02

G

ener

ic p

aram

eter

s for

the

plan

t are

com

pose

d or

sele

cted

. Inf

orm

atio

n fr

om d

iffer

ent

plan

ts is

inte

grat

ed to

obt

ain

suita

ble

gene

ric p

aram

eter

s usi

ng B

ayes

ian

met

hods

, cla

ssic

al

appr

oach

es a

nd e

xper

t jud

gmen

t.

The

sele

ctio

n of

gen

eric

par

amet

ers o

r the

com

posi

tion

of g

ener

ic in

form

atio

n in

to a

pa

ram

eter

app

licab

le fo

r the

pla

nt in

clud

es a

n ap

prec

iatio

n an

d an

ass

essm

ent o

f the

un

certa

intie

s inv

olve

d in

the

orig

inal

dat

a an

d in

usi

ng th

em fo

r the

pla

nt.

86

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

Sp

ecia

l Attr

ibut

e D

A-C

02-S

1:

For n

ew e

quip

men

t, th

e us

e of

gen

eric

dat

a an

d m

anuf

actu

rer d

ata

for

the

asse

ssm

ent o

f rel

iabi

lity

para

met

ers i

s jus

tifie

d.

CO

MM

ENT:

The

use

of o

nly

man

ufac

ture

r dat

a fo

r new

equ

ipm

ent m

ay

not r

efle

ct th

e eq

uipm

ent r

elia

bilit

y in

real

ope

ratio

nal c

ondi

tions

. The

use

of

just

ified

gen

eric

dat

a w

ith su

pple

men

tal c

onsi

dera

tion

of m

anuf

actu

rer

data

may

hel

p to

avo

id e

xces

sive

opt

imis

m in

est

imat

ion

of re

liabi

lity

para

met

ers.

87

T

able

9.2

-D

Att

ribu

tes f

or D

ata

Ana

lysi

s: T

ask

DA

-D ‘

Plan

t-Sp

ecifi

c D

ata

Col

lect

ion

and

Eva

luat

ion’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

DA

-D

Plan

t-spe

cific

dat

a co

llect

ion

and

eval

uatio

n is

per

form

ed in

a c

onsi

sten

t and

syst

emat

ic w

ay.

Plan

t-spe

cific

dat

a is

col

lect

ed in

acc

orda

nce

with

the

para

met

er d

efin

ition

s of T

ask

DA

-A

(Tab

le 9

.2-A

) and

the

grou

ping

ratio

nale

of T

ask

DA

-B (T

able

9.2

-B).

DA

-D01

Pl

ant-s

peci

fic d

ata

for t

he b

asic

eve

nt/p

aram

eter

pop

ulat

ion

corr

espo

ndin

g to

that

def

ined

in

Task

s DA

-A (T

able

9.2

-A) a

nd D

A-B

(Tab

le 9

.2-B

) is c

olle

cted

. The

follo

win

g da

taba

ses a

re

crea

ted:

(1) ‘

Failu

re d

atab

ase’

con

tain

ing

failu

res,

plan

ned

and

unpl

anne

d m

aint

enan

ce a

nd te

st

even

ts.

(2) ‘

Succ

ess d

atab

ase’

con

tain

ing

expo

sure

to d

eman

ds a

nd, d

epen

ding

on

the

type

of

equi

pmen

t, th

e ex

posu

re to

stan

dby

and

oper

atio

n.

Thes

e da

taba

ses a

re li

nked

to th

e eq

uipm

ent/b

asic

eve

nt d

atab

ase.

DA

-D02

Pl

ant-s

peci

fic d

ata

from

as b

road

a ti

me

perio

d as

pos

sibl

e is

col

lect

ed, c

onsi

sten

t with

un

iform

ity in

des

ign,

ope

ratio

nal p

ract

ices

, and

exp

erie

nce.

The

ratio

nale

for s

cree

ning

or

disr

egar

ding

pla

nt-s

peci

fic d

ata

is ju

stifi

ed (e

.g. p

lant

des

ign

mod

ifica

tions

, cha

nges

in

oper

atin

g pr

actic

es).

DA

-D03

W

hen

eval

uatin

g m

aint

enan

ce o

r oth

er re

leva

nt re

cord

s to

extra

ct p

lant

-spe

cific

com

pone

nt

failu

re e

vent

dat

a, a

cle

ar b

asis

for t

he id

entif

icat

ion

of e

vent

s as f

ailu

res i

s req

uire

d:

(a) T

he d

istin

ctio

n is

mad

e be

twee

n th

ose

degr

aded

stat

es fo

r whi

ch fa

ilure

, as m

odel

led

in

the

PSA

, wou

ld h

ave

occu

rred

on

dem

and

(e.g

. an

oper

ator

dis

cove

rs th

at a

pum

p ha

s no

oil i

n its

lubr

icat

ion

rese

rvoi

r), a

nd th

ose

that

wou

ld n

ot (e

.g. s

low

pic

k-up

to ra

ted

spee

d).

(b) A

ll ev

ents

that

wou

ld h

ave

resu

lted

in a

failu

re to

per

form

the

mis

sion

as d

efin

ed in

the

PSA

are

incl

uded

as f

ailu

res.

CO

MM

ENT:

Fai

lure

s in

post

-mai

nten

ance

test

ing

need

to b

e sc

reen

ed

on w

heth

er th

e fa

ilure

is c

ause

d by

the

mai

nten

ance

or d

ue a

pre

-exi

stin

g ca

use.

DA

-D04

Th

e ‘s

ucce

ss d

atab

ase’

for s

tand

by c

ompo

nent

s in

term

s of t

he n

umbe

r of p

lant

-spe

cific

de

man

ds is

det

erm

ined

on

the

basi

s of t

he n

umbe

r of s

urve

illan

ce te

sts,

mai

nten

ance

act

s, su

rvei

llanc

e te

sts o

r mai

nten

ance

on

othe

r com

pone

nts,

and

oper

atio

nal d

eman

ds. A

dditi

onal

de

man

ds fr

om p

ost-m

aint

enan

ce te

stin

g ar

e no

t cou

nted

; tha

t is p

art o

f the

succ

essf

ul

rene

wal

.

Onl

y th

ose

test

s and

test

ing

step

s are

cou

nted

whi

ch re

alis

tical

ly te

st th

e fu

nctio

n of

pa

rticu

lar e

quip

men

t and

whi

ch a

re a

ble

to d

etec

t the

ass

ocia

ted

failu

res m

odes

as a

ppea

ring

in th

e PS

A.

CO

MM

ENT:

Dem

ands

resu

lting

from

pos

t mai

nten

ance

test

ing

do n

ot

need

to b

e in

clud

ed, u

nles

s the

y re

veal

new

failu

res u

nrel

ated

to th

e or

igin

al c

ause

of m

aint

enan

ce.

88

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

DA

-D05

Th

e nu

mbe

r of s

urve

illan

ce te

sts a

nd p

lann

ed m

aint

enan

ce a

ctiv

ities

is b

ased

on

plan

t re

quire

men

ts.

RA

TIO

NA

LE:

Mos

t of t

he d

eman

ds fo

r sta

ndby

com

pone

nts r

esul

t fr

om ro

utin

e ac

tiviti

es, w

hich

are

pre

dict

able

. A

ctua

l dem

ands

are

ge

nera

lly a

min

or a

djus

tmen

t, w

hich

do

not h

ave

a si

gnifi

cant

impa

ct o

n th

e pa

ram

eter

est

imat

es.

Sp

ecia

l Attr

ibut

e D

A-D

05-S

1:

The

num

ber o

f sur

veill

ance

test

s and

pla

nned

mai

nten

ance

act

iviti

es is

ba

sed

on a

ctua

l act

iviti

es a

nd o

pera

tion.

RA

TIO

NAL

E : F

or c

erta

in a

pplic

atio

ns a

real

istic

repr

esen

tatio

n of

th

ese

feat

ures

is re

quir

ed, f

or e

xam

ple

for c

onsi

deri

ng p

reve

ntiv

e m

aint

enan

ce fo

r ope

ratio

nal e

quip

men

t suc

h as

mai

n fe

edw

ater

pum

ps.

DA

-D06

Th

e ‘s

ucce

ss d

atab

ase’

is e

stim

ated

in te

rms o

f ope

ratio

nal t

ime

from

surv

eilla

nce

test

pr

actic

es fo

r sta

ndby

com

pone

nts,

and

from

act

ual o

pera

tiona

l dat

a fo

r nor

mal

ly o

pera

ting

com

pone

nts.

Com

pone

nt o

pera

ting

times

are

est

imat

ed in

term

s of a

ctua

l ope

ratin

g tim

es a

nd p

ract

ices

and

op

erat

ing

times

in su

rvei

llanc

e te

sts.

For n

orm

ally

ope

ratin

g co

mpo

nent

s, re

gula

r sw

itcho

vers

ar

e ta

ken

into

acc

ount

bet

wee

n re

dund

ant c

ompo

nent

s and

trai

ns, a

nd a

ssoc

iate

d te

sts,

whi

ch

are

usua

lly c

arrie

d ou

t at t

he ti

me

of s

witc

hove

rs. E

quip

men

t run

hou

r met

er a

nd st

art c

ount

er

data

are

use

d if

avai

labl

e, e

.g. f

or su

mp

pum

ps.

CO

MM

ENT:

the

run

times

for s

tand

by c

ompo

nent

s in

perio

dica

l su

rvei

llanc

e te

sts m

ay b

e sm

all c

ompa

red

to th

e m

issi

on ti

me

spec

ified

in

the

PSA

. In

this

cas

e, th

e su

rvei

llanc

e te

sts d

o no

t pro

vide

info

rmat

ion

for t

he ti

me

span

bet

wee

n th

e en

d of

the

test

runs

and

the

end

of th

e m

issi

on ti

me.

In p

rinci

ple,

the

beha

viou

r of t

he p

artic

ular

com

pone

nt is

th

en u

ntes

ted

for t

his t

ime

span

whe

n de

man

ded

afte

r an

initi

atin

g ev

ent.

This

situ

atio

n ne

eds s

peci

al c

onsi

dera

tion,

e.g

. the

use

of a

dditi

onal

in

form

atio

n (g

ener

ic, e

quip

men

t man

ufac

ture

r) to

cov

er th

e ex

tend

ed

time

perio

d.

DA

-D07

W

hen

usin

g da

ta o

n m

aint

enan

ce a

nd te

stin

g du

ratio

ns to

est

imat

e un

avai

labi

litie

s at t

he

com

pone

nt, t

rain

, or s

yste

m le

vel,

as re

quire

d by

the

syst

em m

odel

, onl

y th

e un

avai

labi

litie

s fr

om th

ose

mai

nten

ance

or t

est a

ctiv

ities

that

wou

ld le

ave

the

com

pone

nt, t

rain

, or s

yste

m

unab

le to

per

form

its f

unct

ion

whe

n de

man

ded

are

incl

uded

in th

e da

ta se

t.

DA

-D08

W

hen

an u

nava

ilabi

lity

of a

fron

t lin

e sy

stem

com

pone

nt is

cau

sed

by a

n un

avai

labi

lity

of a

su

ppor

t sys

tem

, the

una

vaila

bilit

y is

cou

nted

tow

ards

that

of t

he su

ppor

t sys

tem

and

not

the

fron

t lin

e sy

stem

.

CO

MM

ENT:

Cou

ntin

g th

e un

avai

labi

lity

for t

he fr

ont l

ine

syst

em w

ould

le

ad to

an

over

estim

atio

n of

its u

nava

ilabi

lity.

DA

-D09

Fo

r equ

ipm

ent o

utag

e, th

e du

ratio

n of

the

actu

al ti

me

that

the

equi

pmen

t was

una

vaila

ble

is

iden

tifie

d an

d ev

alua

ted

for e

ach

cont

ribut

ing

activ

ity. S

ince

mai

nten

ance

out

ages

are

a

func

tion

of th

e pl

ant s

tatu

s, on

ly th

ose

outa

ges a

re c

ount

ed w

hich

occ

urre

d du

ring

the

parti

cula

r pla

nt st

atus

for w

hich

mai

nten

ance

una

vaila

bilit

y da

ta is

col

lect

ed.

Spec

ial a

ttent

ion

is p

aid

to th

e ca

se o

f a m

ulti-

plan

t site

with

shar

ed sy

stem

s, w

hen

the

tech

nica

l spe

cific

atio

ns c

an b

e di

ffere

nt d

epen

ding

on

the

stat

us o

f bot

h pl

ants

, whi

ch in

turn

re

quire

s a c

orre

spon

ding

allo

catio

n of

out

age

data

am

ong

basi

c ev

ents

.

DA

-D10

C

oinc

iden

t out

age

times

for r

edun

dant

equ

ipm

ent (

both

intra

- and

inte

r-sy

stem

) are

iden

tifie

d an

d ev

alua

ted

base

d on

act

ual p

lant

exp

erie

nce.

89

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

DA

-D11

Pl

ant-s

peci

fic re

pair

even

ts o

r rel

ated

and

app

licab

le in

dust

ry e

xper

ienc

e ar

e id

entif

ied

and

eval

uate

d fo

r eac

h re

pair

incl

udin

g th

e as

soci

ated

repa

ir tim

e. T

he re

pair

time

is th

e tim

e sp

an

from

the

iden

tific

atio

n of

the

com

pone

nt fa

ilure

unt

il th

e co

mpo

nent

is re

turn

ed to

serv

ice.

DA

-D12

D

ata

on re

cove

ry fr

om lo

ss o

f off

site

pow

er, l

oss o

f ser

vice

of s

ervi

ce w

ater

, etc

. are

rare

on

a pl

ant s

peci

fic b

asis

. If a

vaila

ble,

for e

ach

reco

very

, the

ass

ocia

ted

reco

very

tim

e is

iden

tifie

d an

d ev

alua

ted.

The

reco

very

tim

e is

the

time

span

from

the

iden

tific

atio

n of

the

syst

em o

r fu

nctio

n fa

ilure

unt

il th

e sy

stem

or f

unct

ion

is re

turn

ed to

serv

ice.

EXA

MPL

E: F

or th

e lo

ss o

f off

site

pow

er sp

ecia

l app

roac

hes t

o m

ake

use

of th

e st

atis

tical

dat

a fr

om th

e en

tire

elec

trica

l net

wor

k to

whi

ch th

e pl

ant i

s con

nect

ed h

ave

been

dev

elop

ed (s

ee fo

r exa

mpl

e R

ef. [

16])

.

90

T

able

9.2

-E A

ttri

bute

s for

Dat

a A

naly

sis:

Tas

k D

A-E

‘Der

ivat

ion

of P

lant

-Spe

cific

Par

amet

ers,

Inte

grat

ion

of G

ener

ic a

nd P

lant

-Spe

cific

In

form

atio

n’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

DA

-E

The

para

met

er e

stim

ates

are

bas

ed o

n re

leva

nt g

ener

ic in

dust

ry a

nd p

lant

-spe

cific

evi

denc

e.

Whe

re a

pplic

able

, gen

eric

and

pla

nt-s

peci

fic e

vide

nce

is in

tegr

ated

usi

ng a

ccep

tabl

e m

etho

ds

to o

btai

n pl

ant-s

peci

fic p

aram

eter

est

imat

es. E

ach

para

met

er e

stim

ate

is a

ccom

pani

ed b

y a

char

acte

rizat

ion

of th

e un

certa

inty

.

CO

MM

ENT:

For

man

y ap

plic

atio

ns, g

ener

ic p

aram

eter

est

imat

es m

ay

be a

dequ

ate.

In

gene

ral,

if th

e ge

neric

est

imat

es a

re c

hose

n ca

refu

lly,

and

are

appr

opria

te fo

r the

pla

nt sp

ecifi

c co

mpo

nent

des

ign,

com

pone

nt

boun

dary

and

failu

re m

ode

defin

ition

s, an

d op

erat

iona

l con

ditio

ns, t

he

plan

t spe

cific

est

imat

es w

ould

not

be

expe

cted

to b

e si

gnifi

cant

ly

diffe

rent

. H

owev

er, t

he u

se o

f pla

nt sp

ecifi

c da

ta w

ill re

sult

in a

hig

her

leve

l of c

onfid

ence

in th

e re

sults

of t

he P

SA.

DA

-E01

Pl

ant-s

peci

fic p

aram

eter

est

imat

es a

re c

alcu

late

d us

ing

Bay

esia

n up

date

s whe

re fe

asib

le.

Prio

r dis

tribu

tions

are

sele

cted

as e

ither

non

-info

rmat

ive,

or r

epre

sent

ativ

e of

var

iabi

lity

in

indu

stry

dat

a.

CO

MM

ENT:

Con

stan

t fai

lure

rate

is u

sual

ly p

ostu

late

d as

sum

ing

abse

nce

of a

ging

and

‘chi

ld d

isea

se’ e

ffec

ts.

Sp

ecia

l Attr

ibut

e D

A-E0

1-S1

: A

time

tren

d an

alys

is is

per

form

ed to

exp

lore

the

exis

ting

tren

ds in

the

relia

bilit

y pa

ram

eter

s, in

par

ticul

ar fo

r pas

sive

and

non

-rep

lace

able

co

mpo

nent

s tak

ing

into

acc

ount

the

agin

g ef

fect

s. .

CO

MM

ENT:

The

tim

e tr

end

anal

ysis

is u

sefu

l for

the

appl

icat

ions

de

alin

g w

ith e

xplo

ratio

n of

agi

ng p

heno

men

a.

DA

-E02

If

nei

ther

pla

nt-s

peci

fic d

ata

nor g

ener

ic p

aram

eter

est

imat

es a

re a

vaila

ble

for t

he p

aram

eter

as

soci

ated

with

a sp

ecifi

c ba

sic

even

t, es

timat

es fo

r the

mos

t sim

ilar e

quip

men

t ava

ilabl

e ar

e us

ed, a

djus

ting,

if n

eces

sary

, to

acco

unt f

or d

iffer

ence

s. A

ltern

ativ

ely,

use

can

be

mad

e of

ex

pert

judg

men

t or a

naly

tical

mod

els a

nd th

e ra

tiona

le b

ehin

d th

e ch

oice

of p

aram

eter

val

ues

is d

ocum

ente

d.

DA

-E03

A

mea

n va

lue

of, a

nd a

stat

istic

al re

pres

enta

tion

of th

e un

certa

inty

inte

rval

s for

the

para

met

er

estim

ates

is p

rovi

ded.

C

OM

MEN

T: A

ccep

tabl

e sy

stem

atic

met

hods

incl

ude

for i

nsta

nce:

B

ayes

ian

upda

ting

or e

xper

t jud

gmen

t.

DA

-E04

W

hen

the

Bay

esia

n ap

proa

ch is

use

d to

der

ive

a di

strib

utio

n an

d m

ean

valu

e of

a p

aram

eter

, a

chec

k is

mad

e to

asc

erta

in th

at th

e po

ster

ior d

istri

butio

n de

rived

is re

ason

able

giv

en th

e pr

ior

dist

ribut

ion

and

the

plan

t spe

cific

evi

denc

e.

CO

MM

ENT:

If th

e es

timat

or fo

r the

mea

n va

lue

of a

par

amet

er b

ased

on

pla

nt e

vide

nce

is o

utsi

de o

f a 9

5% c

onfid

ence

inte

rval

aro

und

the

med

ian

valu

e of

the

prio

r dis

tribu

tion

the

appl

icab

ility

of t

hat p

artic

ular

pr

ior d

ata

and

dist

ribut

ion

shou

ld b

e re

cons

ider

ed re

gard

ing

its

appl

icab

ility

to th

e co

mpo

nent

and

failu

re m

ode

unde

r con

side

ratio

n.

91

T

able

9.2

-F

Att

ribu

tes f

or D

ata

Ana

lysi

s: T

ask

DA

-F ‘D

eriv

atio

n of

Pla

nt-S

peci

fic P

aram

eter

s for

Com

mon

Cau

se F

ailu

re E

vent

s’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

DA

-F

Der

ivat

ion

of p

lant

-spe

cific

par

amet

ers f

or C

CF

even

ts. A

s for

the

para

met

er e

stim

ates

for

inde

pend

ent e

vent

s, th

ese

para

met

ers a

re b

ased

on

rele

vant

gen

eric

indu

stry

and

pla

nt-

spec

ific

evid

ence

if a

vaila

ble.

Eac

h pa

ram

eter

est

imat

e is

acc

ompa

nied

by

a ch

arac

teriz

atio

n of

the

unce

rtain

ty.

CO

MM

ENT:

Bec

ause

CC

F ev

ents

are

rare

eve

nts,

only

ver

y fe

w p

lant

sp

ecifi

c da

ta a

re u

sual

ly a

vaila

ble.

The

refo

re C

CF

para

met

er e

stim

atio

n re

lies m

uch

mor

e on

gen

eric

dat

a an

d on

exp

ert j

udgm

ent.

DA

-F01

Th

e B

eta-

fact

or a

ppro

ach

or a

n eq

uiva

lent

met

hod

is u

sed

for C

CF

mod

els a

nd re

gard

ing

the

estim

atio

n of

CC

F pa

ram

eter

s.

Sp

ecia

l Attr

ibut

e D

A-F0

1-S1

: Th

e Be

ta-fa

ctor

app

roac

h or

an

equi

vale

nt m

etho

d is

onl

y us

ed fo

r an

in

itial

scr

eeni

ng s

tep

with

scr

eeni

ng v

alue

s fo

r th

e pa

ram

eter

s. Fo

r th

e fin

aliz

ed m

odel

, a m

odel

is u

sed

whi

ch a

llow

s a

refin

ed r

epre

sent

atio

n of

failu

re c

ombi

natio

ns fo

r hig

her o

rder

redu

ndan

cies

.

RATI

ON

ALE:

Use

of t

he B

eta-

fact

or m

odel

may

mas

ks-o

ut in

term

edia

te

failu

re c

ombi

natio

ns fo

r equ

ipm

ent w

ith m

ore

than

two

redu

ndan

cies

. For

su

ch e

quip

men

t, th

e Be

ta-fa

ctor

mod

el d

oes n

ot p

rovi

de a

real

istic

de

scri

ptio

n of

CC

F ev

ents

.

EXAM

PLE

of m

ore

deta

iled

mod

ellin

g:

(a) A

lpha

Fac

tor M

odel

(b

) Bas

ic P

aram

eter

Mod

el

(c) M

ultip

le G

reek

Let

ter M

odel

(d

) Bin

omia

l Fai

lure

Rat

e M

odel

DA

-F02

G

ener

ic c

omm

on c

ause

failu

re p

roba

bilit

ies a

re u

sed.

R

ATI

ON

ALE

: Si

nce

CC

F ev

ents

are

rare

, unc

erta

intie

s in

estim

ates

of

CC

F pr

obab

ilitie

s are

rela

tivel

y la

rge.

For

man

y ap

plic

atio

ns, t

he u

se o

f ge

neric

par

amet

er v

alue

s is a

ccep

tabl

e, a

lthou

gh th

e co

nclu

sion

s fro

m th

e PS

A n

eed

to b

e as

sess

ed fo

r the

ir ro

bust

ness

with

resp

ect t

o th

e un

certa

inty

in th

e C

CF

prob

abili

ties.

Sp

ecia

l Attr

ibut

e D

A-F0

2-S1

:

Real

istic

com

mon

cau

se fa

ilure

pro

babi

litie

s con

sist

ent w

ith p

lant

-sp

ecifi

c de

sign

and

ope

ratio

nal p

ract

ices

, sup

port

ed b

y pl

ant-s

peci

fic

scre

enin

g an

d m

appi

ng o

f ind

ustr

y-w

ide

data

is u

sed

for d

omin

ant

com

mon

-cau

se e

vent

s.

RATI

ON

ALE:

CC

F ev

ents

usu

ally

hav

e a

maj

or im

pact

on

resu

lts. U

se o

f ge

neri

c C

CF

mod

el p

aram

eter

s may

mas

k-ou

t diff

eren

ces w

hich

ch

arac

teri

ze a

pplic

atio

ns.

CO

MM

ENT:

An

exam

ple

appr

oach

is p

rovi

ded

in N

URE

G/C

R-54

85 (s

ee

Ref.

[13]

). Al

tern

ativ

ely,

an

appr

oach

like

the

Part

ial B

eta-

fact

or is

use

d w

hich

is, h

owev

er, l

imite

d to

the

Beta

-fact

or C

CF

mod

el.

92

T

able

9.2

-G

Att

ribu

tes f

or D

ata

Ana

lysi

s: T

ask

DA

-G ‘H

andl

ing

of R

elia

bilit

y Pa

ram

eter

s Aff

ecte

d by

Pla

nt C

hang

es, H

andl

ing

of N

ew

Equ

ipm

ent’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

DA

-G

Plan

t cha

nges

may

aff

ect a

vaila

ble

relia

bilit

y pa

ram

eter

s by

chan

ging

ope

ratio

nal

cond

ition

s.

For n

ewly

add

ed e

quip

men

t, ne

w p

aram

eter

s are

requ

ired.

As f

or th

e ‘b

ase

case

PSA

’ dat

a an

alys

is, t

he p

aram

eter

est

imat

es a

re b

ased

on

rele

vant

gen

eric

indu

stry

and

pla

nt-s

peci

fic

evid

ence

[see

Tas

ks D

A-B

(Tab

le 9

.2-B

), D

A-C

(Tab

le 9

.2-A

C),

and

DA

-D (T

able

9.2

-A)]

. W

here

app

licab

le, g

ener

ic a

nd p

lant

-spe

cific

evi

denc

e ar

e in

tegr

ated

usi

ng a

ccep

tabl

e m

etho

ds to

obt

ain

plan

t-spe

cific

par

amet

er e

stim

ates

. Eac

h pa

ram

eter

est

imat

e is

ac

com

pani

ed b

y a

char

acte

rizat

ion

of th

e un

certa

inty

.

DA

-G01

If

mod

ifica

tions

to p

lant

des

ign

or o

pera

ting

prac

tice

lead

to a

con

ditio

n w

here

pas

t dat

a ar

e no

long

er re

pres

enta

tive

of c

urre

nt p

erfo

rman

ce li

mit

the

use

of o

ld d

ata:

a)

If th

e m

odifi

catio

n in

volv

es n

ew e

quip

men

t or a

pra

ctic

e w

here

sign

ifica

nt g

ener

ic

para

met

er e

stim

ates

are

ava

ilabl

e, p

aram

eter

est

imat

es u

pdat

ed w

ith p

lant

-spe

cific

dat

a as

it b

ecom

es a

vaila

ble

are

used

, or;

b)

If th

e m

odifi

catio

n is

uni

que

to th

e ex

tent

that

gen

eric

par

amet

er e

stim

ates

are

not

av

aila

ble

and

only

lim

ited

expe

rienc

e is

ava

ilabl

e fo

llow

ing

the

chan

ge, t

hen

the

impa

ct o

f the

cha

nge

is a

naly

sed

and

the

hypo

thet

ical

eff

ect o

n th

e hi

stor

ical

dat

a is

as

sess

ed to

det

erm

ine

to w

hat e

xten

t the

dat

a ca

n be

use

d (s

ee a

lso

Tabl

e 9.

2-E,

G

ener

al A

ttrib

ute

DA

-E02

).

93

T

able

9.2

-H

Att

ribu

tes f

or D

ata

Ana

lysi

s: T

ask

DA

-H ‘D

ocum

enta

tion’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

DA

-H

Doc

umen

tatio

n an

d in

form

atio

n st

orag

e is

per

form

ed in

a m

anne

r tha

t fac

ilita

tes p

eer r

evie

w,

as w

ell a

s fut

ure

upgr

ades

and

app

licat

ions

of t

he P

SA b

y de

scrib

ing

the

proc

esse

s tha

t wer

e us

ed a

nd p

rovi

ding

det

ails

of t

he a

ssum

ptio

ns m

ade

and

thei

r bas

es.

DA

-H01

Th

e fo

llow

ing

is d

ocum

ente

d:

a) S

yste

m a

nd c

ompo

nent

bou

ndar

ies u

sed

to e

stab

lish

com

pone

nt fa

ilure

pro

babi

litie

s;

b) G

roup

ing

crite

ria fo

r com

pone

nts

c) T

he m

odel

use

d to

eva

luat

e ea

ch b

asic

eve

nt p

roba

bilit

y;

d) S

ourc

es fo

r gen

eric

par

amet

er e

stim

ates

; e)

The

pla

nt-s

peci

fic so

urce

s of d

ata;

f)

The

tim

e pe

riods

for w

hich

pla

nt-s

peci

fic d

ata

wer

e ga

ther

ed;

g) K

ey a

ssum

ptio

ns m

ade

in th

e in

terp

reta

tion

of d

ata

and

the

reas

onin

g (b

ased

on

engi

neer

ing

judg

men

t, sy

stem

s mod

ellin

g, o

pera

tions

, and

stat

istic

al k

now

ledg

e)

supp

ortin

g its

use

in p

aram

eter

est

imat

ion;

h)

Jus

tific

atio

n fo

r exc

lusi

on o

f any

dat

a;

i) T

he b

asis

for t

he e

stim

ates

of c

omm

on c

ause

failu

re p

roba

bilit

ies,

incl

udin

g ju

stifi

catio

n fo

r scr

eeni

ng o

r map

ping

of g

ener

ic a

nd p

lant

-spe

cific

dat

a;

j) T

he ra

tiona

le fo

r any

dis

tribu

tions

use

d as

prio

rs fo

r Bay

esia

n up

date

s, w

here

app

licab

le;

k) P

aram

eter

est

imat

es in

clud

ing

the

char

acte

rizat

ion

of u

ncer

tain

ty a

s app

ropr

iate

.

DA

-H02

Th

e de

rivat

ion

of th

e pa

ram

eter

val

ues i

s doc

umen

ted.

The

info

rmat

ion/

data

base

is

docu

men

ted

and

stor

ed in

a w

ay, w

hich

allo

ws t

he re

prod

uctio

n of

the

data

ana

lysi

s tas

k fo

r ex

ampl

e fo

r rel

iabi

lity

para

met

er u

pdat

es.

Sp

ecia

l Attr

ibut

e D

A-H

02-S

1:

The

info

rmat

ion

from

the

data

ana

lysi

s inc

ludi

ng th

e da

taba

ses c

reat

ed is

pa

rt o

f the

PSA

mod

el a

nd d

ocum

enta

tion.

Thi

s dat

a in

clud

ing

the

data

base

s and

the

deta

iled

back

grou

nd in

form

atio

n is

stor

ed in

a

retr

ieva

ble

and

acce

ssib

le e

lect

roni

c fo

rm a

nd fo

rmat

. Due

to th

e am

ount

of

info

rmat

ion

aris

ing

from

the

data

ana

lysi

s tas

ks e

lect

roni

c st

orag

e of

th

is in

form

atio

n is

ess

entia

l for

man

y of

the

appl

icat

ions

.

RATI

ON

ALE:

Cer

tain

app

licat

ions

cou

ld b

e pr

actic

ally

pre

vent

ed

with

out t

he in

form

atio

n be

ing

avai

labl

e an

d ac

cess

ible

in th

is w

ay.

94

10. PSA ELEMENT ‘DF’: DEPENDENT FAILURES ANALYSIS

10.1. Main objectives The objective of the dependent failure analysis is to support other PSA elements with

dependent failure information and assure that all possible dependencies are correctly considered. Correctly modelling dependencies is essential to the development of the PSA model. The dependent failure analysis tasks provides the vehicle for confirming that all dependencies, including subtle dependencies, are included in the PSA, either by explicit modelling or by common cause failure modelling. Dependent failures to be considered and analysed are:

- Design related dependencies - Operational related dependencies - Physical dependencies - Initiator related dependencies - Residual dependencies (common cause failures)

It should be noted that the dependency analysis is exercised as part of almost all other

PSA elements. The categories of dependencies and PSA elements in which they are addressed are provided in Table 10.1. Dependencies that arise from area events (e.g. internal fires and floods) and external initiating events (e.g. earthquakes, high winds) are not included here since they are outside the scope of this publication.

Table 10.1 Dependence Categories and PSA Elements

Dependence Type

Dependence Category Analysis Procedure or Method PSA Elements

Functional dependencies

Development of accident sequences

Development of success criteria

Accident sequence analysis

Success criteria

Systems analysis

Support system dependencies

Development of system models Systems analysis

Accident sequence analysis (large event tree approach) D

esig

n R

elat

ed

Shared component dependencies

Development of system models Systems analysis

Ope

ratio

ns

Rel

ated

Human action dependencies


Human reliability analysis

Modelling of support state initiating events

Initiating events analysis


Systems analysis

Human reliability analysis

95

Dependence Type

Dependence Category Analysis Procedure or Method PSA Elements

Common environmental effects


Development of system models


Systems analysis

Phys

ical

Dynamic effects Physical analyses Initiating events analysis (pipe breaks)

Initi

ator

Common cause initiating events

Analysis of operating experience, insights from other PSA studies, link from functional dependencies

Use of fault tree models

Initiating event analysis

Systems analysis

Res

idua

l D

epen

denc

ies Common cause

failures Definition of CCCGs

Use of accepted CCF model

Systems analysis

Data analysis

10.2. Dependent failure analysis tasks and their attributes

Table 10.2 lists the main tasks for the PSA element ‘Dependent Failure Analysis’.

Tables 10.2-A through 10.2-G present the description of general and special attributes for these tasks.

Table 10.2 Main Tasks for Dependent Failure Analysis


DF-A Design Related Dependency Analysis

DF-B Operations Related Dependency Analysis

DF-C Physical Dependency Analysis

DF-D Common Cause Initiating Event Analysis

DF-E Common Cause Failure Analysis

DF-F Subtle Interactions

DF-G Documentation

96

T

able

10.

2-A

Att

ribu

tes f

or D

epen

dent

Fai

lure

Ana

lysi

s: T

ask

DF-

A ‘

Des

ign

Rel

ated

Dep

ende

ncy

Ana

lysi

s’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

DF-

A

The

desi

gn re

late

d de

pend

enci

es a

re in

clud

ed in

the

acci

dent

sequ

ence

and

syst

em m

odel

s.

DF-

A01

Fu

nctio

nal d

epen

denc

ies a

re a

ddre

ssed

in th

e st

ruct

ure

of th

e ac

cide

nt se

quen

ces m

odel

s.

EXA

MPL

ES:

Func

tiona

l dep

ende

nce

- (B

WR

): if

depr

essu

rizat

ion

of

the

reac

tor f

ails

, the

low

pre

ssur

e in

ject

ion

func

tion

is g

uara

ntee

d to

fail.

(PW

R):

if lo

w p

ress

ure

inje

ctio

n fa

ils, t

he re

circ

ulat

ion

func

tion

fails

.

DF-

A02

Su

ppor

t sys

tem

dep

ende

ncie

s are

mod

elle

d by

link

ing

syst

em m

odel

s (e.

g. fa

ult t

rees

) th

roug

h ap

prop

riate

tran

sfer

gat

es (f

ault

tree

linki

ng a

ppro

ach)

, or b

y de

velo

ping

mod

els f

or

each

supp

ort s

yste

m st

ate

(larg

e ev

ent t

ree/

supp

ort s

tate

/eve

nt tr

ee li

nkin

g ap

proa

ch).

CO

MM

ENT:

The

cor

rect

iden

tific

atio

n of

impl

icit

syst

em d

epen

denc

ies

(not

evi

dent

from

sche

mat

ics)

is c

ruci

al. T

hus,

depe

nden

ce o

n am

bien

t en

viro

nmen

tal c

ondi

tions

is a

lso

a de

sign

dep

ende

nce.

EXA

MPL

E: T

he d

epen

denc

e on

room

tem

pera

ture

and

thus

on

the

vent

ilatio

n sy

stem

and

on

the

room

hea

ting

syst

em

DF-

A03

C

omm

on c

ompo

nent

dep

ende

ncie

s are

mod

elle

d in

the

faul

t tre

es (f

ault

tree

linki

ng

appr

oach

) or b

y ex

plic

itly

incl

udin

g th

e co

mpo

nent

as a

n ev

ent t

ree

bran

ch p

oint

(lar

ge

even

t tre

e/su

ppor

t sta

te/e

vent

tree

link

ing

appr

oach

).

97

T

able

10.

2-B

Att

ribu

tes f

or D

epen

dent

Fai

lure

Ana

lysi

s: T

ask

DF-

B ‘O

pera

tions

Rel

ated

Dep

ende

ncy

Ana

lysi

s’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

DF-

B

Ope

ratio

nal r

elat

ed d

epen

denc

ies a

re a

ddre

ssed

in th

e st

ruct

ure

of th

e ac

cide

nt se

quen

ce

and

syst

em m

odel

s by

the

incl

usio

n of

hum

an fa

ilure

eve

nts.

RA

TIO

NA

LE:

This

cla

ss o

f dep

ende

ncy

addr

esse

s the

impa

ct o

f hum

an

actio

ns o

n th

e su

cces

s or f

ailu

re o

f fun

ctio

ns o

r sys

tem

s, an

d in

clud

es

both

pre

-initi

atin

g ev

ent a

ctio

ns, a

nd a

ctio

ns in

resp

onse

to c

hang

es in

pl

ant c

ondi

tions

(pre

curs

ors t

o in

itiat

ing

even

ts re

sulti

ng fr

om lo

ss o

f su

ppor

t sys

tem

s, an

d in

itiat

ing

even

ts).

CO

MM

ENT:

The

ana

lysi

s of o

pera

tiona

l rel

ated

dep

ende

ncie

s is

prim

arily

add

ress

ed in

the

Initi

atin

g Ev

ent A

naly

sis,

Acc

iden

t Seq

uenc

e A

naly

sis,

Syst

ems A

naly

sis,

and

Hum

an R

elia

bilit

y A

naly

sis.

DF-

B01

Pr

e-in

itiat

or o

pera

tiona

l rel

ated

dep

ende

ncie

s, i.e

. tho

se re

sulti

ng fr

om

test

/mai

nten

ance

/cal

ibra

tion

erro

rs (p

re-in

itiat

or e

rror

s) a

re in

clud

ed in

the

appr

opria

te

syst

em m

odel

s. (S

ee a

lso

Task

s HR

-A th

roug

h H

R-C

[Tab

les 8

.2-A

thro

ugh

8.2-

C])

.

DF-

B02

Th

e de

pend

enci

es a

risin

g fr

om th

e ne

ed fo

r ope

rato

r int

erve

ntio

n in

the

case

of p

artia

l su

ppor

t sys

tem

failu

res (

e.g.

loss

of a

trai

n of

com

pone

nt c

oolin

g w

ater

syst

em) a

s cal

led

for b

y em

erge

ncy

oper

atin

g pr

oced

ures

are

add

ress

ed in

the

syst

em m

odel

s use

d to

eva

luat

e th

e in

itiat

ing

even

t fre

quen

cy.

DF-

B03

D

epen

denc

y of

func

tions

or s

yste

ms o

n op

erat

or in

terv

entio

n fo

llow

ing

an in

itiat

ing

even

t is

mod

elle

d in

the

stru

ctur

e of

the

acci

dent

sequ

ence

and

syst

em m

odel

s (se

e H

R-E

and

H

R-F

).

DF-

B04

Th

e va

lues

of t

he H

EPs u

sed

for t

he H

FEs a

re a

ppro

pria

te fo

r the

pla

nt-s

peci

fic a

nd

scen

ario

- spe

cific

con

ditio

ns. (

See

also

Tas

ks H

R-D

(Tab

le 8

.2-D

) and

HR

-G (T

able

8.2

-G

)).

RA

TIO

NA

LE:

The

perf

orm

ance

of t

he o

pera

tors

, in

parti

cula

r for

re

spon

se a

ctio

ns, i

s con

ditio

ned

by a

num

ber o

f fac

tors

that

are

dep

ende

nt

on th

e sc

enar

io c

hara

cter

istic

s.

DF-

B05

Th

e de

pend

ence

bet

wee

n H

EPs a

ppea

ring

in th

e sa

me

cuts

et is

ass

esse

d an

d th

e va

lues

ch

ange

d as

app

ropr

iate

. (Se

e al

so T

able

8.2

-D, G

ener

al A

ttrib

ute

HR

-D04

, and

Tab

le 8

.2-

G, G

ener

al A

ttrib

ute

HR

-G07

).

98

T

able

10.

2-C

Att

ribu

tes f

or D

epen

dent

Fai

lure

Ana

lysi

s: T

ask

DF-

C: ‘

Phys

ical

Dep

ende

ncy

Ana

lysi

s’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

DF-

C

Dep

ende

ncie

s tha

t aris

e be

caus

e of

deg

rada

tion

of th

e en

viro

nmen

t of t

he p

lant

equ

ipm

ent

are

addr

esse

d. T

hese

incl

ude

dyna

mic

effe

cts o

f a p

ipe

brea

k an

d se

cond

ary

effe

cts o

f ot

her i

nitia

ting

even

ts a

nd e

vent

s occ

urrin

g du

ring

an a

ccid

ent s

cena

rio th

at c

an c

ause

fa

ilure

s of s

afet

y re

late

d eq

uipm

ent,

whi

ch is

nee

ded

for t

he m

itiga

tion

of th

e in

itiat

ing

even

t or f

ailu

res o

f whi

ch c

an m

ake

the

initi

atin

g ev

ent w

orse

.

DF-

C01

Th

e an

alys

is sc

ope

for t

he p

hysi

cal d

epen

denc

ies i

nclu

des t

he fo

llow

ing

cate

gorie

s of

phen

omen

a (s

ee a

lso

Tabl

e 5.

2-C

, Gen

eral

Attr

ibut

e A

S-C

09, a

nd T

able

7.2

-C, G

ener

al

Attr

ibut

es S

Y-C

09, S

Y-C

10, S

Y-C

11):

Cat

egor

y 1

Impa

cts o

f pip

e w

hip,

pro

ject

iles,

and

wat

er/s

team

jets

in c

ase

of p

ipe

brea

ks,

vess

el ru

ptur

es, e

tc.;

the

influ

ence

s can

be

dire

cted

to a

djac

ent e

quip

men

t or

build

ing

stru

ctur

es.

Cat

egor

y 2

Con

sequ

ence

s of i

ncre

ased

hum

idity

and

tem

pera

ture

.

Cat

egor

y 3

D

istri

butio

n of

isol

atio

n w

ool m

ater

ial a

nd b

lock

ing

of th

e re

circ

ulat

ion

flow

(p

ost –

LOC

A).

RA

TIO

NA

LE:

Phys

ical

effe

cts c

an su

bsta

ntia

lly re

duce

the

miti

gatio

n po

ssib

ilitie

s.

99

T

able

10.

2-D

Att

ribu

tes f

or D

epen

dent

Fai

lure

Ana

lysi

s: T

ask

DF-

D ‘

Com

mon

Cau

se In

itiat

ing

Eve

nt A

naly

sis’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

DF-

D

An

anal

ysis

of c

omm

on c

ause

initi

atin

g ev

ents

is p

erfo

rmed

.

DF-

D01

C

omm

on c

ause

initi

atin

g ev

ents

are

iden

tifie

d an

d in

clud

ed in

the

PSA

mod

el a

s nec

essa

ry.

(See

als

o Ta

ble

4.2-

A, G

ener

al A

ttrib

utes

IE-A

04, I

E-A

06, a

nd T

able

4.2

-E, G

ener

al

Attr

ibut

e IE

-E03

).

CO

MM

ENTS

: A

com

mon

cau

se in

itiat

or (C

CI)

is a

n ev

ent c

ausi

ng a

tra

nsie

nt (o

r req

uirin

g m

anua

l shu

tdow

n) a

nd a

t the

sam

e tim

e de

grad

ing

one

or m

ore

safe

ty fu

nctio

ns th

at m

ay b

e ne

eded

afte

r the

tran

sien

t/shu

t-do

wn.

CC

Is a

re re

stric

ted

to fa

ilure

s occ

urrin

g in

side

the

plan

t sys

tem

s, su

ch a

s fa

ilure

s in

the

cont

rol a

nd p

rote

ctio

ns sy

stem

s, el

ectri

c po

wer

supp

ly

syst

em, s

ervi

ce w

ater

syst

em o

r oth

er su

ppor

t sys

tem

s.

The

CC

I ana

lysi

s is u

sual

ly a

par

t of t

he In

itiat

ing

Even

ts D

efin

ition

and

G

roup

ing

Task

and

the

Syst

ems A

naly

sis T

ask,

exc

ept f

or th

e m

odel

ling

of th

e pl

ant r

espo

nse

to a

n id

entif

ied

and

impo

rtant

CC

I, w

hich

bel

ongs

to

the

Acc

iden

t Seq

uenc

e A

naly

sis.

100

T

able

10.

2-E

Att

ribu

tes f

or D

epen

dent

Fai

lure

Ana

lysi

s: T

ask

DF-

E ‘

Com

mon

Cau

se F

ailu

re A

naly

sis’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

DF-

E

A c

ompo

nent

com

mon

cau

se fa

ilure

ana

lysi

s (C

CF)

is p

erfo

rmed

.

DF-

E01

Com

mon

cau

se fa

ilure

s are

incl

uded

in th

e sy

stem

mod

els a

s app

ropr

iate

(see

Tas

ks S

Y-C

in

Tab

le 7

.2-C

and

DA

-F in

Tab

le 9

.2-F

). R

ATI

ON

ALE

: W

hile

man

y so

urce

s of d

epen

denc

y ca

n be

mod

elle

d ex

plic

itly

in th

e PS

A m

odel

, the

re a

re fa

ilure

mec

hani

sms a

t the

co

mpo

nent

leve

l tha

t can

resu

lt in

mul

tiple

star

t fai

lure

s whe

n co

mpo

nent

s are

dem

ande

d or

mul

tiple

com

pone

nt fa

ilure

s with

in th

e m

issi

on ti

me

requ

ired

of th

ose

com

pone

nts.

Thes

e ar

e th

e so

-cal

led

‘com

mon

cau

se fa

ilure

s’.

101

T

able

10.

2-F

Att

ribu

tes f

or D

epen

dent

Fai

lure

Ana

lysi

s: T

ask

DF-

F ‘S

ubtle

Inte

ract

ions

’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

DF-

F A

n an

alys

is o

f sub

tle in

tera

ctio

n de

pend

enci

es is

per

form

ed.

DF-

F01

Ope

ratio

nal e

xper

ienc

e of

the

plan

t bei

ng a

naly

sed

and

sim

ilar p

lant

s is r

evie

wed

to

iden

tify

even

ts th

at in

volv

e un

usua

l dep

ende

nt e

ffec

ts. P

SA m

odel

add

ress

es th

ese

effe

cts.

Spec

ial A

ttrib

ute

DF-

F01-

S1:

His

tori

cal e

vent

s tha

t inv

olve

unu

sual

dep

ende

nt e

ffect

s are

revi

ewed

to

dete

rmin

e w

heth

er su

ch o

ccur

renc

es c

an o

ccur

at t

he p

lant

bei

ng

anal

ysed

.

RATI

ON

ALE:

A st

ruct

ured

iden

tific

atio

n an

d co

nsid

erat

ion

of su

btle

in

tera

ctio

ns m

ay b

e ne

eded

for c

ompl

eten

ess p

urpo

ses.

CO

MM

ENT:

So

calle

d su

btle

dep

ende

ncie

s5 are

not

ord

inar

y fu

nctio

nal

depe

nden

cies

but

are

spec

ific

to a

ctua

l dem

and

cond

ition

s, w

hen

the

plan

t sys

tem

s are

act

uate

d an

d op

erat

ed u

nder

tran

sien

t or e

mer

genc

y co

nditi

ons.

Typi

cally

, sub

tle d

epen

denc

ies a

re n

ot d

etec

ted

in n

orm

al o

pera

tion

or b

y su

rvei

llanc

e te

sts.

The

inte

ract

ion

betw

een

syst

ems o

r sub

syst

ems c

an b

e tr

ansm

itted

by

the

proc

ess m

ediu

m, v

ia su

ppor

t sys

tem

rout

es o

r in

dire

ctly

via

ope

ratin

g en

viro

nmen

t, e.

g. te

mpe

ratu

re, h

umid

ity, p

ress

ure

wav

es o

r vib

ratio

n.

Subt

le d

epen

denc

ies a

re e

ither

func

tiona

l or p

hysi

cal,

but t

hey

are

diffi

cult

to fo

rese

e. Id

entif

ied

subt

le d

epen

denc

ies c

an b

e tr

eate

d by

ex

plic

it m

odel

ling,

or c

onsi

dere

d in

the

CC

F m

odel

s.

EXAM

PLES

: A

list o

f sub

tle in

tera

ctio

ns c

an b

e fo

und

in N

URE

G-1

150

(see

Ref

. [17

]).

5 Sub

tle d

epen

denc

ies a

re a

lso

calle

d ‘s

yste

m in

tera

ctio

ns’ o

r ‘su

btle

inte

ract

ions

’.

102

T

able

10.

2-G

Att

ribu

tes f

or D

epen

dent

Fai

lure

Ana

lysi

s: T

ask

DF-

G ‘

Doc

umen

tatio

n’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

DF-

G

Doc

umen

tatio

n an

d in

form

atio

n st

orag

e is

per

form

ed i

n a

man

ner

that

fac

ilita

tes

peer

re

view

, as

wel

l as

futu

re u

pgra

des

and

appl

icat

ions

of t

he P

SA b

y de

scrib

ing

the

proc

esse

s th

at w

ere

used

and

pro

vidi

ng d

etai

ls o

f the

ass

umpt

ions

mad

e an

d th

eir b

ases

.

DF-

G01

Fu

nctio

nal d

epen

denc

ies a

re d

ocum

ente

d in

the

form

of c

ompo

nent

info

rmat

ion

tabl

es a

nd

syst

em d

epen

denc

y m

atric

es in

a c

lear

and

trac

eabl

e m

anne

r. C

OM

MEN

T: It

is u

sefu

l to

prod

uce

an in

tegr

ated

dep

ende

ncy

mat

rix to

pr

ovid

e an

ove

rvie

w o

f the

func

tiona

l dep

ende

ncie

s ove

r all

syst

ems.

Sp

ecia

l Attr

ibut

e D

F-G

01-S

1:

The

info

rmat

ion

cont

aine

d in

the

com

pone

nt in

form

atio

n ta

bles

, pr

otec

tion

sign

al ta

bles

, dep

ende

ncy

mat

rice

s etc

. is s

tore

d in

a

rela

tiona

l dat

abas

e.

CO

MM

ENT:

Doc

umen

tatio

n of

com

plex

and

inte

rrel

ated

dat

a in

the

form

of

a re

latio

n da

taba

se is

impo

rtan

t for

app

licat

ions

bas

ed o

n Li

ving

PSA

(R

isk

Mon

itor)

.

DF-

G02

Th

e sp

ecifi

c as

sum

ptio

ns a

nd li

mita

tions

con

cern

ing

func

tiona

l dep

ende

ncie

s are

do

cum

ente

d.

EXA

MPL

ES: T

he fo

llow

ing

item

s are

exa

mpl

es o

f gen

eric

ass

umpt

ions

an

d lim

itatio

ns:

- In

som

e pl

ant r

oom

s, th

e fa

ilure

of r

oom

coo

ling/

heat

ing

can

cons

titut

e a

CC

I. Si

mila

rly, s

peci

fic fa

ilure

situ

atio

ns in

oth

er su

ppor

t sy

stem

s can

lead

to C

CIs

, whi

ch a

re a

naly

sed

sepa

rate

ly.

- Fa

ilure

of c

ompo

nent

pro

tect

ion

is n

ot c

onsi

dere

d as

failu

re if

it is

lik

ely

that

the

com

pone

nt w

ill su

rviv

e th

e de

man

d an

d ne

eded

m

issi

on ti

me

(will

fulfi

l the

safe

ty fu

nctio

n ev

en th

ough

deg

rade

d).

DF-

G03

Fo

r the

CC

I ana

lysi

s, th

e fo

llow

ing

is d

ocum

ente

d:

- D

escr

iptio

n of

the

CC

I ide

ntifi

catio

n pr

oces

s as d

efin

ed in

the

Initi

atin

g Ev

ents

D

efin

ition

and

Gro

upin

g Ta

sk

- Li

st o

f the

CC

Is c

onsi

dere

d as

initi

atin

g ev

ents

with

a re

fere

nce

whe

re th

e as

soci

ated

ac

cide

nt se

quen

ce a

naly

sis a

nd e

vent

tree

mod

ellin

g ar

e de

scrib

ed

- C

CIs

that

wer

e sc

reen

ed o

ut.

Com

plex

cas

es a

re d

iscu

ssed

in th

e in

itiat

ing

even

t ana

lysi

s or i

n th

e co

rres

pond

ing

syst

ems

anal

ysis

repo

rt, in

ord

er to

exp

lain

the

deta

ils o

f the

cau

sal m

odel

ling,

spec

ial i

nitia

tor

char

acte

ristic

s and

der

ivat

ion

of th

e in

itiat

or fr

eque

ncy.

DF-

G04

Fo

r the

phy

sica

l dep

ende

ncy

anal

ysis

, the

follo

win

g is

doc

umen

ted:

- Ex

tens

ions

to L

OC

A a

nd tr

ansi

ent c

ateg

orie

s -

Ref

inem

ents

to a

ccid

ent s

eque

nce

mod

els o

f LO

CA

s and

tran

sien

ts

- Ev

alua

tion

of c

onta

inm

ent s

ump

oper

abili

ty.

103

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

DF-

G05

Fo

r the

CC

F an

alys

is, t

he fo

llow

ing

is d

ocum

ente

d:

- Id

entif

icat

ion

and

defin

ition

of C

CC

Gs

- A

ll C

CC

Gs a

nd c

ompo

nent

s inc

lude

d

- Tr

eatm

ent o

f int

ra-s

yste

m a

nd in

ters

yste

m C

CFs

-

Mod

els u

sed

- Sp

ecia

l CC

F ev

ents

-

CC

F da

ta

- A

des

crip

tion

of h

ow C

CF

is im

plem

ente

d in

the

over

all P

SA m

odel

.

104

11. PSA ELEMENT ‘MQ’: MODEL INTEGRATION AND CDF QUANTIFICATION


The main objective of the model integration and quantification process is to develop an integrated plant-specific PSA model that will be used to estimate the core damage frequency and to develop an understanding of the contributors to core damage model (see also Section 12). The following principles should be met:

• The PSA model reflects the current design, operational practices (procedures,

configuration strategy), and the operational experience.

• The parameter uncertainties are considered in the model.

• The quantification is performed correctly, taking into account the dependencies discussed in Section 10.

• The results are reviewed to ensure that the solution reflects the plant characteristics.

• Illogical or incorrect minimal cutsets are removed.

• Recovery actions are applied to minimal cutsets that are assessed to be too conservative, as needed.

11.2. Model integration and CDF quantification tasks and their attributes

Table 11.2 lists the main tasks for the PSA element ‘Model Integration and CDF

Quantification’. Tables 11.2-A through 11.2-D provide the description of general and special attributes for these tasks.

Table 11.2 Main Tasks for Model Integration and CDF Quantification


MQ-A Integrated Model

MQ-B Requirements on the Quantification

MQ-C Review and Modification of the Results

MQ-D Documentation

105

T

able

11.

2-A

Att

ribu

tes f

or M

odel

Inte

grat

ion

and

CD

F Q

uant

ifica

tion:

Tas

k M

Q-A

‘In

tegr

ated

Mod

el’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

MQ

-A

Con

stru

ctio

n of

an

inte

grat

ed p

lant

-spe

cific

PSA

mod

el fr

om th

e el

emen

ts a

ddre

ssed

in

Sect

ions

4 th

roug

h 10

is p

erfo

rmed

.

MQ

-A01

A

ll sy

stem

mod

els a

nd a

ccid

ent s

eque

nces

mod

els a

re in

tegr

ated

to p

rovi

de th

e lo

gic

stru

ctur

e of

the

PSA

mod

el.

For t

he fa

ult t

ree

linki

ng a

ppro

ach,

at e

ach

bran

ch p

oint

, the

cor

resp

ondi

ng fa

ult t

ree

is

solv

ed fo

r the

func

tion

or sy

stem

succ

ess c

riter

ia a

nd b

ound

ary

cond

ition

s tha

t ref

lect

the

scen

ario

spec

ific

plan

t con

ditio

ns.

Sup

port

syst

em d

epen

denc

ies a

re a

ddre

ssed

by

the

cons

iste

nt e

vent

-nam

ing

sche

me

used

for f

ault

tree

cons

truct

ion

and

esta

blis

hing

logi

c lin

ks

betw

een

the

syst

em m

odel

s (fr

ontli

ne-to

-sup

port

syst

ems,

supp

ort-t

o-su

ppor

t sys

tem

s).

For t

he e

vent

tree

link

ing

appr

oach

, the

pro

babi

litie

s of t

he b

ranc

h po

ints

(som

etim

es c

alle

d sp

lit fr

actio

ns) a

re c

ondi

tiona

l on

the

path

thro

ugh

the

even

t tre

e.

In b

oth

case

s, th

e co

nditi

ons i

nclu

de th

e im

pact

of t

he su

cces

s or f

ailu

re o

f fun

ctio

ns o

r sy

stem

s req

uire

d pr

ior t

o th

e fu

nctio

n or

syst

em o

f con

cern

, whi

ch h

as a

n im

pact

on

the

succ

ess c

riter

ia, b

oth

in te

rms o

f the

num

ber o

f tra

ins r

equi

red

and

the

timin

g fo

r req

uire

d fu

nctio

ns o

r sys

tem

s, w

hich

in tu

rn a

ffect

s the

hum

an re

liabi

lity

anal

ysis

app

ropr

iate

to th

e ev

ent.

RA

TIO

NA

LE:

Ther

e ar

e tw

o co

mm

only

use

d ap

proa

ches

to P

SA lo

gic

mod

el c

onst

ruct

ion:

the

so-c

alle

d fa

ult t

ree

linki

ng a

ppro

ach;

and

the

linke

d ev

ent t

ree

appr

oach

. Th

e fo

rmer

relie

s on

com

pute

r cod

es th

at u

se

Boo

lean

logi

c to

add

ress

dep

ende

ncie

s aris

ing

from

com

mon

com

pone

nts

or c

omm

on su

ppor

t sys

tem

s. F

or th

ese

mod

els,

the

use

of a

con

sist

ent

nam

ing

sche

me

for t

he e

vent

s and

gat

es in

the

faul

t tre

es is

ess

entia

l.

The

latte

r dep

ends

on

the

cons

iste

nt a

pplic

atio

n of

logi

c ru

les t

hat i

dent

ify

the

appr

opria

te c

ondi

tions

for t

he so

lutio

n of

sys

tem

mod

els s

o th

at th

e ev

ents

on

the

even

t tre

e ca

n be

trea

ted

as in

depe

nden

t.

CO

MM

ENT:

Ini

tiatin

g ev

ents

for w

hich

acc

iden

t seq

uenc

e ar

e no

t de

velo

ped,

but

that

are

ass

umed

to le

ad d

irect

ly to

cor

e da

mag

e sh

ould

al

so b

e in

clud

ed in

the

inte

grat

ed m

odel

. Ex

ampl

es th

at a

re so

met

imes

un

deve

lope

d in

clud

e IS

LOC

A, a

nd re

acto

r ves

sel r

uptu

re.

MQ

-A02

Fo

r the

faul

t tre

e lin

king

app

roac

h, lo

gic

loop

s are

cut

in a

man

ner t

hat m

inim

izes

loss

of

info

rmat

ion

and

does

not

pro

duce

ung

roun

ded

optim

istic

resu

lts b

y, fo

r exa

mpl

e, re

mov

ing

som

e of

the

depe

nden

cies

(by

negl

ectin

g pa

rts o

f the

supp

ort s

yste

m lo

gic

failu

re m

odel

).

RA

TIO

NA

LE: W

hen

faul

t tre

es a

re li

nked

to c

reat

e an

inte

grat

ed lo

gic

mod

el, l

ogic

loop

s can

app

ear d

ue to

mut

ual d

epen

denc

ies b

etw

een

supp

ort s

yste

ms,

e.g.

see

belo

w.

EXA

MPL

E: T

he c

lass

ical

exa

mpl

e is

des

crib

ed in

NU

REG

/CR

-272

8 (R

ef. [

18])

dea

ling

with

an

emer

genc

y di

esel

gen

erat

or d

epen

ding

on

serv

ice

wat

er fo

r coo

ling

whi

ch in

turn

requ

ires e

lect

ric p

ower

from

the

dies

el. M

any

cont

empo

rary

PSA

s hav

e re

fined

and

ext

ende

d m

odel

s for

el

ectri

cal p

ower

supp

lies a

nd c

ontro

l I&

C w

hich

may

cre

ate

mor

e co

mpl

ex lo

gic

loop

s. Th

ese

logi

c lo

ops h

ave

to b

e re

solv

ed, b

asic

ally

by

cutti

ng th

e lo

op a

t an

appr

opria

te p

lace

or l

evel

bec

ause

, oth

erw

ise,

the

mod

el c

anno

t be

reso

lved

and

qua

ntifi

ed. T

he c

uttin

g of

logi

c lo

ops i

s ca

rrie

d ou

t in

a m

anne

r, w

hich

min

imiz

es th

e lo

ss o

f inf

orm

atio

n an

d do

es

not p

rodu

ce n

on-c

onse

rvat

ive

resu

lts. I

t is c

arrie

d ou

t acc

ordi

ng to

a

defin

ed c

once

pt a

nd p

roce

dure

and

the

deta

ils o

f res

olvi

ng lo

gic

loop

s are

fu

lly d

ocum

ente

d.

106

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

MQ

-A03

Th

e in

itiat

ing

even

t fre

quen

cies

and

the

prob

abili

ties a

ssoc

iate

d w

ith b

asic

eve

nts o

f the

m

odel

are

con

sist

ent w

ith th

e de

finiti

ons o

f the

eve

nts i

n th

e co

ntex

t of t

he lo

gic

mod

el.

C

OM

MEN

T: T

he p

aram

eter

s are

def

ined

in su

ch a

way

that

they

re

pres

ent t

he p

lant

spec

ific

desi

gn a

nd o

pera

tiona

l exp

erie

nce

as w

ell a

s sc

enar

io sp

ecifi

c bo

unda

ry c

ondi

tions

(esp

ecia

lly im

porta

nt fo

r hum

an

erro

r pro

babi

litie

s) a

s dis

cuss

ed in

Sec

tions

4 -

10.

107

T

able

11.

2-B

Att

ribu

tes f

or M

odel

Inte

grat

ion

and

CD

F Q

uant

ifica

tion:

Tas

k M

Q-B

‘R

equi

rem

ents

on

the

Qua

ntifi

catio

n’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

MQ

-B

Solu

tion

of th

e m

odel

to d

eriv

e th

e lo

gica

l com

bina

tions

of e

vent

s lea

ding

to c

ore

dam

age

(e.g

. min

imal

cut

sets

) and

the

quan

tific

atio

n of

the

core

dam

age

freq

uenc

y is

per

form

ed

with

an

appr

opria

te c

ode.

CO

MM

ENT:

Thi

s pub

licat

ion

refe

rs to

min

imal

cut

sets

as a

n im

porta

nt

resu

lt of

PSA

. Thi

s ter

m a

pplie

s to

the

faul

t tre

e te

chni

que

for P

SA m

odel

s as

sum

ing

cohe

rent

logi

c m

odel

s. Fo

r non

-coh

eren

t log

ic m

odel

s, th

e co

rres

pond

ing

mat

hem

atic

al te

rm is

‘prim

e im

plic

ants

’.

EXA

MPL

E: T

he p

rior g

ener

atio

n of

cut

sets

is n

ot re

quire

d fo

r the

qu

antif

icat

ion.

Prim

e im

plic

ants

can

be

deriv

ed o

n th

e ba

sis o

f the

ge

nera

tion

of B

inar

y D

ecis

ion

Dia

gram

s (B

DD

s). B

DD

s are

a sp

ecifi

c re

pres

enta

tion

of S

truct

ural

Boo

lean

Log

ic F

unct

ions

of c

ompl

ex s

yste

ms

allo

win

g th

e ex

act q

uant

ifica

tion

of fa

ult t

rees

incl

udin

g no

n-co

here

nt

logi

c m

odel

s (m

odel

s usi

ng n

egat

ive

logi

c, i.

e. N

OT-

Gat

es).

MQ

-B01

Th

e co

mpu

ter c

ode

used

for s

olut

ion

and

quan

tific

atio

n of

the

PSA

mod

el is

ver

ified

and

va

lidat

ed, a

nd is

use

d on

ly w

ithin

its s

peci

fied

rang

e of

app

licab

ility

. Spe

cific

lim

itatio

ns o

f th

e co

de a

re re

cogn

ized

.

EXA

MPL

E fo

r spe

cific

lim

itatio

ns th

at m

ay o

ccur

in a

link

ed fa

ult t

ree

mod

el: I

f the

eve

nt tr

ee q

uant

ifica

tion

uses

a su

cces

s pro

babi

lity

of 1

, si

gnifi

cant

qua

ntifi

catio

n er

rors

can

occ

ur, i

f the

failu

re p

roba

bilit

y of

a

linke

d fa

ult t

ree

is h

igh.

MQ

-B02

Th

e PS

A m

odel

is so

lved

and

qua

ntifi

ed to

allo

w id

entif

icat

ion

of th

e si

gnifi

cant

sequ

ence

s co

ntrib

utin

g to

cor

e da

mag

e fr

eque

ncy.

MQ

-B03

Fo

r the

faul

t tre

e lin

king

app

roac

h, th

e fin

al tr

unca

tion

valu

e is

just

ified

by

a se

nsiti

vity

an

alys

is d

emon

stra

ting

that

the

core

dam

age

freq

uenc

y do

es n

ot si

gnifi

cant

ly c

hang

e, if

the

trunc

atio

n va

lue

is re

duce

d.

CO

MM

ENT:

A ru

le o

f thu

mb

says

that

the

trunc

atio

n va

lue

shou

ld b

e 3

orde

rs o

f mag

nitu

de lo

wer

than

the

dom

inan

t val

ue (e

.g. t

he C

DF)

that

is

cons

ider

ed. H

owev

er, o

nly

a se

nsiti

vity

stud

y ca

n as

sure

that

an

appr

opria

te tr

unca

tion

valu

e w

as a

pplie

d.

Sp

ecia

l Attr

ibut

e M

Q-B

03-S

1:

For t

he d

eriv

atio

n of

som

e im

port

ance

list

s (es

peci

ally

RAW

) the

tr

unca

tion

valu

e is

re-e

valu

ated

and

just

ified

. (Se

e al

so T

able

12.

2-A,

G

ener

al A

ttrib

ute

RI-A

03).

RATI

ON

ALE:

If th

is is

not

don

e, se

vera

l low

-pro

babi

lity

even

ts m

ight

be

excl

uded

from

the

impo

rtan

ce li

st w

hene

ver R

AW is

sign

ifica

nt fo

r the

ap

plic

atio

n.

Sp

ecia

l Attr

ibut

e M

Q-B

03-S

2:

If th

e ap

plic

atio

n is

rela

ted

to a

spec

ific

initi

atin

g ev

ent,

of a

spec

ific

grou

p of

sequ

ence

s, th

e at

trib

ute

is a

pplie

d fo

r tho

se c

ontr

ibut

ions

to

CD

F ra

ther

than

the

tota

l CD

F.

CO

MM

ENT:

Sen

sitiv

ity te

sts s

houl

d be

con

duct

ed fo

r the

set o

f res

ults

of

inte

rest

.

108

Ta

sk /

GA

D

escr

iptio

n of

Tas

k/G

ener

al A

ttrib

utes

Id

entif

ier a

nd D

escr

iptio

n of

Spe

cial

Attr

ibut

es (i

n Ita

lics)

R

atio

nale

/Com

men

ts/E

xam

ples

for:

Gen

eral

Attr

ibut

es a

nd

S

peci

al A

ttrib

utes

(in

Italic

s)

MQ

-B04

C

ore

dam

age

freq

uenc

y is

eva

luat

ed a

s a m

ean

valu

e an

d th

e un

certa

inty

dis

tribu

tion

char

acte

ristic

s are

pro

vide

d, fo

r the

tota

l CD

F an

d fo

r the

sign

ifica

nt in

divi

dual

acc

iden

t se

quen

ces.

Para

met

er u

ncer

tain

ties a

ssoc

iate

d w

ith H

EPs,

com

pone

nt re

liabi

lity

para

met

ers,

initi

atin

g ev

ent f

requ

enci

es, e

tc. a

re p

ropa

gate

d (v

ia fa

ult a

nd e

vent

tree

s) th

roug

h th

e m

odel

. C

orre

latio

ns b

etw

een

unce

rtain

ty d

istri

butio

ns a

re a

ddre

ssed

and

con

side

red

for t

he

unce

rtain

ty q

uant

ifica

tion.

Whe

n us

ing

a M

onte

Car

lo, o

r oth

er si

mul

atio

n ap

proa

ch, t

he n

umbe

r of s

imul

atio

ns u

sed

has b

een

dem

onst

rate

d to

pro

duce

stab

le re

sults

.

CO

MM

ENT:

A p

oint

est

imat

e ob

tain

ed b

y su

bstit

utin

g a

mea

n va

lue

for

each

par

amet

er in

the

min

imal

cut

set e

quat

ion,

with

out a

pro

paga

tion

of

unce

rtain

ty g

ives

an

appr

oxim

atio

n to

the

mea

n va

lue.

The

clo

sene

ss o

f th

e ap

prox

imat

ion

to th

e tru

e m

ean

valu

e de

pend

s on

the

failu

re e

vent

s in

clud

ed in

the

cuts

ets,

the

exte

nt o

f cor

rela

tion

betw

een

unce

rtain

ty

dist

ribut

ions

and

the

shap

e of

unc

erta

inty

dis

tribu

tions

invo

lved

.

MQ

-B05

W

hen

usin

g lo

gic

flags

(als

o de

sign

ated

as h

ouse

eve

nts)

, the

logi

c fla

g ev

ents

shou

ld e

ither

se

t to

be lo

gica

l tru

e or

fals

e (in

stea

d of

setti

ng th

e ba

sic

even

t pro

babi

lity

to 1

.0 o

r 0.0

), pr

ior t

o th

e qu

antif

icat

ion

of th

e se

quen

ces.

EXA

MPL

E: L

ogic

flag

s may

be

used

to m

odel

gua

rant

eed

succ

ess o

r fa

ilure

, or t

o sw

itch

on o

r off

the

mod

els c

orre

spon

ding

to d

iffer

ent

conf

igur

atio

ns.

109

T

able

11.

2-C

Att

ribu

tes f

or M

odel

Inte

grat

ion

and

CD

F Q

uant

ifica

tion:

Tas

k M

Q-C

‘R

evie

w a

nd M

odifi

catio

n of

the

Res

ults

’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

MQ

-C

The

task

incl

udes

a re

view

of r

esul

ts a

nd c

orre

ctio

ns to

the

inte

grat

ed m

odel

mad

e as

ne

cess

ary.

R

ATI

ON

ALE

: D

epen

ding

on

the

way

the

inte

grat

ed m

odel

has

bee

n de

velo

ped,

som

e po

st-p

roce

ssin

g of

the

resu

lts m

ay b

e ne

eded

as

disc

usse

d be

low

. A

revi

ew o

f the

resu

lts a

cts a

s a c

onfir

mat

ion

that

the

mod

el h

as b

een

inte

grat

ed c

orre

ctly

.

MQ

-C01

A

sam

ple

of si

gnifi

cant

min

imal

cut

sets

or s

eque

nces

of e

ach

even

t tre

e ty

pe is

revi

ewed

in

orde

r to

dete

rmin

e, if

the

logi

c of

the

min

imal

cut

sets

or s

eque

nces

is c

orre

ct. A

s a sp

ot

chec

k, a

sam

ple

of le

ss si

gnifi

cant

cut

sets

or s

eque

nces

is re

view

ed.

Min

imal

cut

sets

(or s

eque

nces

) con

tain

ing

even

ts th

at a

re m

utua

lly e

xclu

sive

but

app

ear

beca

use

of th

e ap

proa

ch to

mod

ellin

g (e

.g. i

f NO

T ga

tes a

re n

ot u

sed

to e

limin

ate

disa

llow

ed m

aint

enan

ce, o

r mul

tiple

initi

ator

s) a

re id

entif

ied

and

corr

ecte

d.

M

Q-C

02

Spec

ial A

ttrib

ute

MQ

-C02

-S1:

The

syst

em m

odel

s are

inco

rpor

ated

such

that

eac

h co

nfig

urat

ion

of a

sy

stem

(e.g

. no

mai

nten

ance

, mai

nten

ance

on

Trai

n A,

mai

nten

ance

on

Trai

n B,

etc

.) is

mod

elle

d se

para

tely

with

an

appr

opri

ate

time

frac

tion.

(S

ee a

lso

Tabl

e 7.

2-B,

Gen

eral

Attr

ibut

es S

Y-B0

2 an

d SY

-B08

).

RATI

ON

ALE:

A m

ore

deta

iled

mod

ellin

g ap

proa

ch re

quir

es le

ss p

ost-

proc

essi

ng, e

.g. i

n a

Risk

Mon

itor a

pplic

atio

n.

MQ

-C03

M

inim

um c

ut se

ts (o

r seq

uenc

es) c

onta

inin

g m

ultip

le o

pera

tor a

ctio

ns a

re id

entif

ied

and

the

degr

ee o

f dep

ende

ncy

asse

ssed

(see

Tab

le 8

.2-D

, Gen

eral

Attr

ibut

e H

R-D

04, a

nd T

able

8.

2-G

, Gen

eral

Attr

ibut

e H

R-G

07).

RA

TIO

NA

LE:

Bec

ause

the

depe

nden

cy b

etw

een

HFE

s is a

ccid

ent

scen

ario

dep

ende

nt, i

t is o

ften

addr

esse

d by

pos

t-pro

cess

ing

to a

djus

t the

co

mbi

ned

prob

abili

ty o

f the

set o

f dep

ende

nt H

FEs.

MQ

-C04

R

ecov

ery

actio

ns a

re in

clud

ed in

the

quan

tific

atio

n pr

oces

s in

appl

icab

le se

quen

ces a

nd

min

imal

cut

sets

. Rec

over

y ac

tions

cre

dite

d in

the

eval

uatio

n ar

e ei

ther

pro

cedu

raliz

ed o

r ha

ve re

ason

able

like

lihoo

d of

succ

ess a

ssum

ing

that

trai

ned

and

qual

ified

per

sonn

el a

re

perf

orm

ing

the

reco

very

act

ion(

s). (

See

also

Tas

k H

R-H

in T

able

8.2

-H).

110

T

able

11.

2-D

Att

ribu

tes f

or M

odel

Inte

grat

ion

and

CD

F Q

uant

ifica

tion:

Tas

k M

Q-D

‘D

ocum

enta

tion’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

MQ

-D

Doc

umen

tatio

n is

per

form

ed in

a m

anne

r tha

t fac

ilita

tes p

eer r

evie

w a

nd fu

ture

upg

rade

s.

The

follo

win

g as

pect

s of t

he m

odel

inte

grat

ion

and

CD

F qu

antif

icat

ion

proc

ess a

re d

ocum

ente

d:

- th

e in

tegr

atio

n pr

oces

s of a

ll fa

ult a

nd e

vent

tree

s;

- PS

A m

odel

ver

sion

;

- a

gene

ral d

escr

iptio

n of

the

quan

tific

atio

n pr

oces

s;

- th

e pr

oces

s and

the

resu

lts fo

r est

ablis

hing

the

trunc

atio

n va

lues

;

- so

ftwar

e ve

rsio

n an

d qu

antif

icat

ion

setu

p (tr

unca

tion

valu

es, e

tc.);

- th

e pr

oces

s and

the

resu

lts o

f the

qua

ntifi

catio

n re

view

;

- th

e m

ean

valu

e an

d th

e un

certa

inty

dis

tribu

tion

of th

e to

tal C

DF,

eac

h cl

ass o

f ini

tiatin

g ev

ents

, and

of s

igni

fican

t cor

e da

mag

e se

quen

ces;

- th

e ac

cide

nt se

quen

ces a

nd th

eir c

ontri

butin

g cu

t set

s.

M

Q-D

01

Spec

ial A

ttrib

ute

MQ

-D01

-S1:

W

hen

only

a fr

actio

n of

the

CD

F is

requ

ired

to b

e an

alys

ed fo

r an

appl

icat

ion,

th

e do

cum

enta

tion

supp

orts

this

ana

lysi

s.

111

12. PSA ELEMENT ‘RI’: RESULTS ANALYSIS AND INTERPRETATION

12.1. Main objectives The objective of the results analysis and interpretation activity is to derive an

understanding of those aspects of plant design and operation that have an impact on the risk. In addition, an important part of this task is to identify the key sources of uncertainty in the model and assess their impact on the results.

Uncertainties can be thought of as being of three main types:

• Parameter uncertainty: these are uncertainties in the values of the initiating event frequencies, component failure probabilities, human error probabilities, etc. These uncertainties can be propagated through the analysis to generate an assessment of the uncertainty on the overall quantitative results using standard methods. Parameter uncertainties are addressed in Section 11.

• Model uncertainty: There are questions with how to model certain failures (e.g. RCP seal LOCAs), or how to represent the impact of plant conditions on system success criteria, for example, for which there is no universally accepted approach. Typically, in PSAs, these model uncertainties are dealt with by making assumptions and adopting a specific model. In relatively rare cases, alternate models may be incorporated into the PSA, weighting each model by a probability representing the degree of belief in that model as being the most appropriate.

• Completeness uncertainty: This is the most difficult to deal with as it represents those contributors to risk that are not included in the model. If the PSA model only includes internal initiating events at power, the contributors to risk not modelled include external events, and alternate modes of operation. At a more subtle level, typically PSAs do not include contributions from errors of commission.

The significance to risk of individual contributors (initiating events, accident

sequences, functional failures, system failures, component failures, human failures, etc.) are explored to derive an understanding of the risk profile of the plant, i.e., what is the impact of various aspects of plant design and operation on risk. The impact of uncertainties and assumptions on the PSA results are addressed in order to determine the robustness of the conclusions concerning the risk profile. The analytical tools used for the analysis of results are importance analyses and sensitivity analyses.

12.2. Results analysis and interpretation tasks and their attributes

Table 12.2 lists the main tasks for the PSA element ‘Results Analysis and

Interpretation’. Tables 12.2-A through 12.2-C present the description of general and special attributes for these tasks.

112

Table 12.2 Results Analysis and Interpretation Tasks


RI-A Identification of Significant Contributors

RI-B Assessment of Assumptions

RI-C Documentation

113

T

able

12.

2-A

Att

ribu

tes f

or R

esul

ts A

naly

sis a

nd In

terp

reta

tion:

Tas

k R

I-A

‘Ide

ntifi

catio

n of

Sig

nific

ant C

ontr

ibut

ors’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

RI-

A

The

task

incl

udes

iden

tific

atio

n of

sign

ifica

nt c

ontri

buto

rs to

the

risk

prof

ile o

f the

pla

nt.

RI-

A01

Si

gnifi

cant

con

tribu

tors

to C

DF

are

iden

tifie

d. T

he c

ontri

buto

rs a

re, i

n in

crea

sing

leve

l of

reso

lutio

n:

- In

itiat

ing

even

ts

- A

ccid

ent s

eque

nces

-

Key

safe

ty fu

nctio

n fa

ilure

s -

Syst

em fa

ilure

s -

Bas

ic e

vent

s

The

basi

c ev

ents

incl

ude:

-

Equi

pmen

t fai

lure

s or u

nava

ilabi

litie

s -

Com

mon

cau

se fa

ilure

s -

Hum

an fa

ilure

eve

nts

RI-

A02

Si

gnifi

cant

con

tribu

tors

to a

ccid

ent s

eque

nces

, key

safe

ty fu

nctio

ns, a

nd sy

stem

s are

id

entif

ied.

R

ATI

ON

ALE

: Rev

iew

s of t

he so

lutio

ns to

syst

em m

odel

s, fu

nctio

nal

mod

els a

nd a

ccid

ent s

eque

nces

are

an

esse

ntia

l par

t of t

he v

alid

atio

n of

th

e st

ruct

ure

of th

e pl

ant l

ogic

mod

el, a

nd fu

rther

mor

e, p

rovi

de a

dditi

onal

in

sigh

ts o

n th

e ris

k pr

ofile

of t

he p

lant

.

RI-

A03

W

hen

asse

ssin

g th

e si

gnifi

canc

e of

bas

ic e

vent

s usi

ng im

porta

nce

mea

sure

s tha

t inv

olve

se

tting

failu

re p

roba

bilit

ies t

o un

ity, s

uch

as th

e ris

k ac

hiev

emen

t wor

th (R

AW

), th

e as

sess

men

t is p

erfo

rmed

by

reso

lvin

g th

e PS

A m

odel

rath

er th

an re

-qua

ntify

ing

a pr

e-so

lved

cut

set l

ist.

RA

TIO

NA

LE:

A c

utse

t lis

t gen

erat

ed w

ith to

o hi

gh tr

unca

tion

valu

e w

ill

excl

ude

som

e co

mpo

nent

s and

ther

efor

e th

eir R

AW

val

ues a

re id

entic

ally

un

ity.

An

alte

rnat

ive

to re

solv

ing

the

mod

el is

to u

se a

cut

set e

quat

ion

solv

ed a

t a lo

wer

trun

catio

n va

lue.

114

T

able

12.

2-B

Att

ribu

tes f

or R

esul

ts A

naly

sis a

nd In

terp

reta

tion:

Tas

k R

I-B

‘Ass

essm

ent o

f Ass

umpt

ions

’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

RI-

B

Ass

essm

ent o

f the

sign

ifica

nce

of k

ey a

ssum

ptio

ns a

nd m

odel

unc

erta

intie

s is p

erfo

rmed

.

RI-

B01

Si

gnifi

cant

sour

ces o

f mod

el u

ncer

tain

ty a

re id

entif

ied

(see

the

com

men

t).

CO

MM

ENT:

A m

odel

unc

erta

inty

is o

ne a

ssoc

iate

d w

ith th

e m

odel

ling

of a

n is

sue

or p

heno

men

on, f

or w

hich

ther

e is

no

cons

ensu

s app

roac

h.

Such

mod

el u

ncer

tain

ties a

re ty

pica

lly a

ddre

ssed

by

adop

ting

one

of a

nu

mbe

r of m

odel

s, or

mak

ing

an a

ssum

ptio

n ab

out t

he im

pact

of a

ph

enom

enon

on

the

oper

abili

ty o

f a sy

stem

or f

unct

ion.

A si

gnifi

cant

so

urce

of u

ncer

tain

ty is

one

whe

re th

e ad

optio

n of

a d

iffer

ent m

odel

or a

di

ffere

nt a

ssum

ptio

n ca

n al

ter t

he si

gnifi

canc

e of

a c

ontri

buto

r. Si

nce

diffe

rent

app

licat

ions

mak

e us

e of

diff

eren

t res

ults

, the

sign

ifica

nt so

urce

s of

unc

erta

inty

will

diff

er b

etw

een

appl

icat

ions

.

RA

TIO

NA

LE:

Whe

n th

e re

sults

of t

he P

SA a

re to

be

used

for d

ecis

ion

mak

ing,

the

deci

sion

-mak

er n

eeds

to b

e aw

are

of th

e im

pact

of t

he

unce

rtain

ties o

n th

e PS

A re

sults

use

d in

the

deci

sion

(see

RI-

B02

).

EXA

MPL

ES o

f pot

entia

l sou

rces

of m

odel

unc

erta

inty

incl

ude:

-

Succ

ess c

riter

ia

- R

CP

seal

LO

CA

mod

el

- A

ssum

ptio

ns a

bout

the

nece

ssity

for r

oom

coo

ling

- C

hoic

e of

qua

ntifi

catio

n m

odel

for h

uman

err

or p

roba

bilit

ies.

RI-

B02

Th

e ef

fect

of a

sign

ifica

nt a

ssum

ptio

n on

the

resu

lts is

ass

esse

d by

per

form

ing

sens

itivi

ty

stud

ies,

usin

g di

ffer

ent p

laus

ible

ass

umpt

ions

.

RA

TIO

NA

LE:

Und

erst

andi

ng o

f the

impa

ct o

f mod

ellin

g un

certa

inty

on

the

resu

lts o

f the

PSA

is c

ruci

al to

a d

ecis

ion-

mak

er.

An

exce

ptio

n ca

n be

w

hen

a pa

rticu

lar m

odel

has

bee

n ap

prov

ed a

s the

stan

dard

mod

el.

CO

MM

ENT:

Som

e se

nsiti

vity

stud

ies c

an b

e pe

rfor

med

by

chan

ges t

o pa

ram

eter

val

ues,

or tu

rnin

g of

f par

ts o

f the

mod

el to

repr

esen

t gro

ups o

f fa

ilure

. O

ther

s may

invo

lve

addi

ng n

ew p

ortio

ns to

the

mod

el.

Thes

e ar

e do

ne in

a m

anne

r con

sist

ent w

ith th

e re

leva

nt a

ttrib

utes

.

RI-

B03

Fo

r mod

el u

ncer

tain

ties o

r ass

umpt

ions

aff

ectin

g th

e sa

me

parts

of t

he P

SA m

odel

, se

nsiti

vity

stud

ies a

re p

erfo

rmed

sim

ulta

neou

sly

to d

eter

min

e w

heth

er th

ere

are

syne

rgis

tic

effe

cts.

RI-

B04

Th

e ch

oice

of t

he sp

ecifi

c as

sum

ptio

ns o

r mod

els a

dopt

ed fo

r the

bas

e ca

se m

odel

is

just

ified

. R

ATI

ON

ALE

: U

nder

stan

ding

the

reas

ons f

or c

hoos

ing

a sp

ecifi

c as

sum

ptio

n or

mod

el is

impo

rtant

for t

he d

ecis

ion-

mak

er.

115

T

able

12.

2-C

Att

ribu

tes f

or R

esul

ts A

naly

sis a

nd In

terp

reta

tion:

Tas

k R

I-C

‘D

ocum

enta

tion’

Task

/ G

A

Des

crip

tion

of T

ask/

Gen

eral

Attr

ibut

es

Iden

tifie

r and

Des

crip

tion

of S

peci

al A

ttrib

utes

(in

Italic

s)

Rat

iona

le/C

omm

ents

/Exa

mpl

es fo

r: G

ener

al A

ttrib

utes

and

Spe

cial

Attr

ibut

es (i

n Ita

lics)

RI-

C

Doc

umen

tatio

n of

resu

lts is

per

form

ed in

a w

ay th

at fa

cilit

ates

und

erst

andi

ng o

f the

te

chni

cal b

asis

for t

he si

gnifi

canc

e of

con

tribu

tors

.

RI-

C01

Th

e re

sults

of t

he a

naly

sis o

f the

sign

ifica

nce

of c

ontri

buto

rs a

re p

rese

nted

in a

var

iety

of

way

s to

char

acte

rize

the

risk

prof

ile o

f the

pla

nt, i

.e.,

wha

t are

the

aspe

cts o

f des

ign

and

oper

atio

nal p

ract

ices

that

hav

e an

impa

ct o

n ris

k, h

ow th

ey im

pact

risk

, and

why

.

RI-

C02

Th

e re

sults

of t

he se

nsiti

vity

ana

lyse

s are

doc

umen

ted

so th

at th

e im

pact

of e

ach

sign

ifica

nt

assu

mpt

ion

is c

hara

cter

ized

app

ropr

iate

ly, a

nd th

e ch

oice

of t

he a

ssum

ptio

n or

mod

el fo

r th

e ba

se c

ase

just

ified

.

116

13. DETERMINATION OF SPECIAL ATTRIBUTES FOR PSA APPLICATIONS As mentioned earlier in the publication, it is assumed that general attributes described

in the publication characterize a contemporary state of the art PSA performed with the aim of assessing the overall NPP safety. The special attributes provide elevated capabilities within particular PSA elements to meet special needs of PSA applications.

In order to indicate what specific features of PSA are needed for particular PSA

applications, Tables 13.1 through 13.6 were developed that provide information in a structured form on which special attributes are appropriate for the applications included in the PSA application categories defined in Section 2 and briefly described in Appendix II. The special attributes in these tables are distinguished as ‘essential’ and ‘supplemental’.

An essential special attribute emphasises a feature of a PSA element that is

considered to be important to generate the results needed to reliably support the PSA application. Failure to meet an essential special attribute may preclude meaningful use of the study for an intended application.

Supplemental special attributes are those that are not necessarily important for a

specific application but could further enhance the usefulness of the PSA by providing a greater level of detail, or improving confidence in the results. In general, it is expected that failure to meet a supplemental attribute does not have a significant impact on the overall results of the PSA application, but may limit the fidelity of the study for certain applications.

In the frame of this publication, it is difficult to foresee all possible PSA application

cases and their specific features. Therefore, the division into essential and supplemental special attributes is to some extent subjective but it is included here to provide a general orientation on the importance of the special attributes in relation to PSA applications. It should be also noted that there are cases when the same special attribute could be considered as essential for some of the applications and supplemental for the others.

Table 13.7 provides a mapping of the special attributes of the PSA elements to the list

of PSA applications presented in Section 2 thus providing an overview of what sets of the special attributes are appropriate for each application.

The following steps provide a practical approach to using the information presented in

this publication to determine the special attributes appropriate for the application of interest:

STEP (1) Identify the PSA application category(s) and specific application(s) from the list presented in Section 2. If needed, consult Appendix II to get supporting information characterizing PSA applications.

STEP (2) Consult Table 13.7 for the set of identifiers of essential and supplemental special

attributes relevant for the application.6 If several applications are planned, the corresponding attributes have to be selected and considered jointly.

6 In Table 13.7, the identifiers of essential special attributes are provided in bold underlined font. The identifiers of supplemental special attributes are provided in regular font.

117

STEP (3) Consult relevant information presented in Tables 13.1-13.6 in order to get understanding of what features of the PSA could be achieved if the special attributes are met.

STEP (4) Consult Section 3.1 to get information on which sections of the publication address

the selected special attributes.7 STEP (5) Consult corresponding Sections 4-12 of the publication to get information on the

content of the special attributes and the features of PSA elements needed to meet them.

7 The first two letters of the identifier of a special attribute indicate the corresponding PSA element.

118

Table 13.1 Safety Assessment

SA Identifier Relevance to Specific Applications in the Category

Essential Attributes

IE-F04-S1 The time trend analysis in relation to IEs provides important information for insights regarding the plant behaviour if changes in the number of specific events are observed at the plant (Application 1.2).

DA-E01-S1 A time trend analysis is important for the assessment of aging effects of the equipment and allows for gaining a realistic risk estimate for operating plant (Application 1.2).

Supplemental Attributes

IE-H02-S1 The availability of information on IEs in the form of an electronic database is useful for periodic PSA updating, which is often part of the periodic safety review (Application 1.2).

IE-A06-S1 Explicit consideration of IEs caused by operator errors, especially those with severe consequences, helps to gain additional insights regarding the plant protection against malevolent acts (Application 1.3).

IE-B01-S1 Reconsideration of the events (especially those with severe consequences), which were screened out based on low frequency in the ‘base case PSA’, helps to gain additional insights regarding the protection of the plant against malevolent acts (Application 1.3).

AS-C02-S1 Plant-specific realistic thermal hydraulic analyses are useful to assure a realistic representation of the specific plant features influencing the accident progression (Application 1.2).

AS-C04-S1 Analysis of the defence against malevolent acts may be enhanced by a thorough consideration of possible mitigation strategies taking into account the impact of malevolent acts for Application 1.3.

AS-C12-S1 Detailed modelling of ATWS sequences provides a better understanding of the impact of reactor protection system failures in the plant risk profile (Application 1.2).

AS-C15-S1 An expanded graphical representation of the accident progression is helpful for documenting the accident sequence models, as well as for understanding, updating, and use of the PSA for Applications 1.2 and 1.3.

SC-B01-S1 The use of proven computer codes and realistic models help to avoid conservative and simplifying success criteria which may mask out the differences or effects of changes (Application 1.2).

HR-G04-S1 The time windows for operator actions that are based on plant-specific thermal-hydraulic analyses provide a more realistic input in the HRA and thus promote getting plant-specific insights dealing with operator performance (Application 1.2).

DA-H02-S1 A periodic safety review requires an update of the reliability parameters, which in turn is facilitated by plant equipment failure data that are electronically retrievable and an evaluation that is retrievable and reproducible (Application 1.2).

DF-F01-S1 A structured identification and consideration of subtle interactions based on historical information from other plants is helpful for the completeness of the PSA model used for realistic estimation of changes in plant risk (Application 1.2).

DF-G01-S1 A relational database containing information on different dependencies and their interconnections is helpful for providing the completeness of the PSA model used for realistic estimation of changes in plant risk (Application 1.2).

119

Table 13.2 Design Evaluation



SY-C07-S1 Realistic estimation of the failure probabilities in specific system conditions provides important information for insights regarding the plant behaviour for newly designed NPPs (Application 2.1) or provides a better understanding of the deviations between an existing plant design and revised design-related rules (Application 2.2).

SY-C14-S1 SY-C15-S1

Consideration of inter-system common cause failures and more detailed modelling of CCF are important for realistic estimation of the plant risk for newly designed NPPs (Application 2.1); it provides also a better understanding of the deviations between an existing plant design and revised design-related rules (Application 2.2).

DA-F01-S1 Decisions on the number of redundancies and diversities to be included in the design require a detailed modelling of CCFs (Applications 2.1 and 2.2).

DA-C02-S1 For newly designed plants, when plant-specific data are not available, the choice of generic data becomes important for justifiable risk results (Application 2.1).


IE-A06-S1 Explicit consideration of IEs caused by operator errors helps to get additional insights regarding deficiencies in the control room design, plant procedures, and operator training. This additional consideration may be used for improvement of the design (Application 2.1).

IE-C02-S1 IE-C02-S2

More detailed grouping of the events at the design phase helps to identify and eliminate deficiencies in the systems design (Application 2.1).

AS-B03-S1 AS-C02-S1 AS-C03-S1 AS-C14-S1

Use of best estimate codes, plant specific t/h analyses, and best estimates of important modelling parameters may help to gain additional insights for evaluation of design features and possible alternatives (Application 2.1).

AS-C12-S1 Detailed modelling of ATWS sequences provides support for the evaluation of the effectiveness of reactor protection system design from the risk perspective (Application 2.1).

AS-C15-S1 An expanded graphical representation of accident progression is helpful for documentation of accident sequence models, as well as for understanding, updating, and use of the PSA (Application 2.1).

SC-B01-S1 The use of proven computer codes and realistic models help to avoid conservative and simplifying success criteria which may mask out differences or effects of changes (Application 2.2).

SY-B13-S1 Revisiting the components screening process promotes a more realistic estimation of the risk for newly designed NPPs at different stages of the design (Application 2.1) and a better understanding of the deviations between an existing plant design and revised design-related rules (Application 2.2).

SY-B15-S1 A detailed modelling of the components promotes a better appreciation of the risk impact of constituting parts of components for newly designed NPPs (Application 2.1) and provides a better understanding of the deviations between an existing plant design and revised design-related rules (Application 2.2).

HR-E01-S2 A search for plant states that could lead to conditions conducive to errors of commission can lead to the identification of defences to prevent such occurrences (Application 2.1).

DF-F01-S1 A structured identification and consideration of subtle interactions is helpful for modelling dependencies for newly designed NPPs (Application 2.1) and assessing the safety importance of deviations between an existing plant design and updated/revised design-related rules (Application 2.2).

120


DF-G01-S1 A relational database containing information on different dependencies is helpful for providing the completeness of the PSA model used for the assessment of safety importance of deviations between an existing plant design and updated/revised design-related rules (Application 2.2).

MQ-B03-S1 MQ-B03-S2 MQ-D01-S1

Use of reduced truncation values for quantification of the whole PSA model or in relation to the particular IEs of major interest may be useful for the assessment of safety importance of deviations between an existing plant design and updated/revised design-related rules (Application 2.2).

121

Table 13.3 NPP Operation



IE-B01-S1 It is important to reconsider the events screened out based on low frequency that are impacted by the equipment subjected to the TS exemption (Application 3.4.3).

IE-F04-S1 Time trend analysis is essential to assess the impact of components aging, which may be the cause for the changes in the number of IEs (Application 3.1.3).


Use of best estimate codes, plant specific t/h analyses, and best estimates of important modelling parameters is essential for adequate modelling of accident scenarios being addressed in the emergency operating procedures for Applications 3.2.1, 3.3.1, and 3.4.2.

AS-C04-S1 AS-C05-S1 AS-C06-S1

Incomplete modelling of accident progression dealing with requirements of plant emergency procedures may cause incomplete or inadequate coverage of accident scenarios for Applications 3.2.1, 3.3.1, and 3.4.2.

SY-B08-S1 The possibility to effectively handle maintenance basic events allows assessment of the impact of maintenance program optimization (Application 3.1.1), impact of improvements in maintenance personnel training program (Application 3.3.2), and is essential for the real time configuration assessment and control, configuration planning, exemptions to TS, justification for continued operation, and dynamic risk-informed TS (Application Group 3.4).

SY-B08-S2 Modelling of test and maintenance unavailabilities at train level is needed for the real time configuration assessment and control, configuration planning, exemptions to TS, justification for continued operation, dynamic risk-informed TS, and maintenance program optimization (Application 3.1.1 and Application Group 3.4).

SY-B08-S3 Symmetric models are needed for Risk Monitor type applications in order to avoid overestimation or underestimation of specific plant configurations (Application Group 3.4).

SY-B13-S1 SY-B15-S1

Maintenance program and ageing management may include more components than those already modelled in the system models and the models would need to be extended to the level of detail needed for the assessment of the impact of maintenance program optimization (Applications 3.1.1 and 3.1.3).

SY-B16-S1 Time-dependent modelling is needed for application dealing with changes to test activities (Application 3.4.1).

HR-G02-S1 The capability of the HRA method used to evaluate the impact of procedure changes is essential for Applications 3.2.1 and 3.2.2.

HR-G04-S1 The time line for the human interactions for dominant ASs defined based on realistic plant-specific thermal-hydraulic analyses is essential for the development and improvement of emergency operating procedures and improvement of operator training programs (Applications 3.2.1 and 3.3.1).

DA-B01-S1 Too broad grouping of components will prevent the identification of specific features of members of the group, which is of importance to maintenance planning and configuration control activities (Applications 3.1.1 and all of Application Group 3.4).

DA-D05-S1 Realistic representation in the model of surveillance tests and planned maintenance activities is essential for maintenance program optimization and configuration planning (Applications 3.1.1 and 3.4.1).

DA-E01-S1 The time trend analysis is essential for activities dealing with optimization of plant aging management program (Application 3.1.3).

MQ-C02-S1 A more detailed modelling of specific system configurations (e.g. maintenance on specific trains) is important for maintenance program optimization (Application 3.1.1) and all Risk Monitor type applications (Application Group 3.4).

122



IE-A06-S1 1) Changes in the maintenance program may impact the probability of operator errors leading to IEs. It is worthwhile to assess the influence of this particular aspect (Application 3.1.1). 2) It is useful to consider the IEs caused by operator errors to avoid adverse impact of changes in maintenance, tests, and training activities (Applications 3.3.1, 3.3.2, and 3.4.1), as well as to identify and assess the events caused by operator errors that may initiate an accident sequence with severe consequences (Applications 3.2.2 and 3.2.3). 3) It is useful to consider IEs caused by human errors while developing management training program (Application 3.3.3).

IE-B01-S1 1) Previously screened IEs caused by equipment failures may appear in the list with a higher frequency due to changes in the maintenance program (Application 3.1.1). 2) It is useful to re-consider low frequent events that have a potential to cause un-mitigated releases (Applications 3.2.2 and 3.2.3).

IE-F04-S1 The time trend analysis is useful in assessing the impact of changes in the maintenance, test, and training activities on the occurrence of IEs (Applications 3.3.1, 3.3.2, 3.3.3, and 3.4.1).

IE-H02-S1 An electronic IE database is helpful in assessing the impact of components aging, which may be the cause for the changes in the number of IEs (Application 3.1.3).


Use of best estimate codes, plant specific t/h analyses, and best estimate of important modelling parameters is useful for realistic modelling of accident scenarios being used for the improvement of management training programs based on plant-specific PSA insights (Application 3.3.3).

AS-C04-S1 AS-C05-S1 AS-C06-S1

A more complete modelling of accident progression dealing with requirements of plant emergency procedures is useful for the improvement of management training programs (Application 3.3.3).

AS-C15-S1 An expanded graphical representation of accident progression is helpful for documentation of accident sequence models, as well as for understanding, updating, and use of the PSA for Applications 3.3.1, 3.3.2, 3.3.3, and 3.4.2.

SY-B08-S1 SY-B08-S2

The possibility to effectively handle test and maintenance basic events may be useful for the improvement of operator training programs (Application 3.3.1).

SY-B13-S1 SY-B15-S1

Improvement of operator training programs may ask for inclusion of more components than those already modelled in the system models (Application 3.3.1).

HR-D06-S1 An identification of past problems can help to identify improvements in maintenance practices (Applications 3.1.1 and 3.3.2).

HR-E01-S1

Modelling of the actions to recover from the equipment failure to automatically initiate or change state may be useful for assessing specific impacts of changes in emergency operating procedures (Application 3.2.1).

HR-E01-S2 Identification of the conditions and plant status conducive to errors of commission could reduce the potential for such errors (Application 3.3.1).

HR-G04-S1 Improvement of the management training program would benefit from a more realistic HRA that use the time windows for operator actions which are based on plant-specific thermal-hydraulic analyses (Application 3.3.3).

DA-D05-S1 Realistic representation in the model of surveillance tests and planned maintenance activities is useful for getting insights for improvement of operator and management training programs (Applications 3.3.1 and 3.3.3).

DA-F01-S1 Detailed modelling of CCFs reflecting the combinations of redundancies/diversities would lead to a more accurate analysis to support plant configuration control (Application Group 3.4).

123


DA-F02-S1 Since CCFs often have a major impact on system unavailability, particularly in highly-redundant systems, a more realistic and specific modelling of CCF events would improve the plant configuration control applications (Application Group 3.4).

DF-F01-S1 DF-G01-S1

A structured identification and consideration of subtle interactions and availability of a relational database containing information on different dependencies is helpful for maintenance program optimization (Application 3.1.1), support for plant ageing management program (Application 3.1.3), development and improvement of the emergency operating procedures and NPP accident management (Applications 3.2.1 and 3.2.2). Availability of a database containing information on different dependencies provides possibility to efficiently maintain a living PSA model for Risk Monitor type applications (Application Group 3.4).

MQ-C02-S1 A more detailed modelling of specific system configurations (e.g. maintenance on specific trains) is helpful for the improvement of operator and management training programs (Applications 3.3.1 and 3.3.3).

124

Table 13.4 Permanent Changes to the Operating Plant



IE-A06-S1 It is important to verify that the changes to TS (e.g. transfer of equipment maintenance from shutdown to power operation) do not impose additional unacceptable risk caused by introducing a potential for new IEs due to operator errors (Applications 4.2.1 and 4.2.2). In addition, an understanding of how human errors during testing contribute to initiating event frequencies and component failures is needed to balance the positive and negative aspects of surveillance testing (Application 4.2.3).

IE-F04-S1 IE-H02-S1

The time trend analysis in relation to IEs and the incidence database are essential for the decisions on NPP upgrades, back-fitting activities, plant modifications, and life-time extension (Application Group 4.1).


Use of best estimate codes, plant specific t/h analyses, and best estimates of important modelling parameters is essential for adequate modelling of the accident scenarios being addressed in the emergency operating procedures (Application 4.1.1).

SC-B01-S1 The use of proven computer codes and realistic models helps to avoid conservative and simplifying success criteria; this is essential for the assessment of the effects of changes (Application 4.1.1).

SY-B08-S1 The possibility to effectively handle maintenance basic events allows assessing the impact of changes to the allowed outage time and required TS actions (Application 4.2.1).

SY-B08-S2 A detailed maintenance model allows an assessment of the impact of changes to the allowed outage time and required TS actions (Application 4.2.1).

SY-B08-S3 Symmetric models are essential for realistic evaluation of changes to AOT, required TS actions, surveillance test intervals, as well as for the equipment risk significance evaluation and evaluation of changes to QA requirements (Applications 4.2.1, 4.2.2, 4.2.3, 4.3.1, 4.3.2).

SY-B16-S1 Time-dependent failure models allow for the assessment of impact of NPP upgrades (Application 4.1.1), changes to the allowed outage time and required TS actions, changes to surveillance test intervals (Applications 4.2.1, 4.2.2, 4.2.3), and risk-informed in-service testing (Application 4.2.4).

SY-C01-S1 PSA applications dealing with optimization of ISI require adequate modelling of the impact of pipe ruptures if the latter are included in the model (Application 4.2.5).

DA-C02-S1 Use of generic sources of data for reliability parameters for new equipment should be justified (Application 4.1.1).

DA-E01-S1 A time trend analysis to explore the existing trends in the reliability parameters is important for Application 4.1.2.

MQ-C02-S1 A more detailed modelling of specific system configurations is important for applications dealing with changes in TS (Application Group 4.2).


IE-A06-S1 Analysis of the IEs caused by operator errors is useful to estimate the impact of changes in the in-service testing or plant modifications (Application 4.1.1 and 4.2.4).

IE-B01-S1 NPP upgrades and changes in in-service testing may increase the frequency of previously screened low frequent events (Applications 4.1.1 and 4.1.2).

IE-C02-S1 IE-C02-S2

Merging of the IE in a single group may mask the impact of plant modifications or changes in testing/inspections activities (Applications 4.1.1, 4.2.4, and 4.2.5).

IE-F04-S1 The time trend analysis in relation to IEs provides helpful information for decision making about equipment risk significance (Applications 4.1.2, 4.3.1, and 4.3.2).

IE-H02-S1 An electronic IE database is useful for Applications 4.1.1, 4.1.2, 4.3.1, and 4.3.2.

AS-C04-S1 Incomplete modelling of accident progression dealing with requirements of plant emergency

125


AS-C05-S1 AS-C06-S1

procedures may cause incomplete or inadequate coverage of accident scenarios for Application 4.1.1.

AS-C15-S1 An expanded graphical representation of accident progression is helpful for documentation of accident sequence models, as well as for understanding, updating, and use of the PSA for Application 4.1.1.

SY-B06-S1

Modelling of pipe failures in the PSA provides the possibility to more precisely assess the importance to risk of particular pipelines/pipe segments that would give a more robust input to the applications dealing with optimization of ISI (Application 4.2.5).

SY-B08-S1 SY-B08-S2

A realistic representation of test and maintenance unavailabilities in system models is useful for risk-informed in-service testing (Application 4.2.4).

SY-B13-S1 Other components than those originally modelled may need to be included in the PSA (Applications 4.1.1, 4.2.1, 4.2.2, 4.2.3, 4.2.5, 4.3.1, 4.3.2).

SY-B15-S1 SY-C15-S1

A deeper level of model resolution down to the level of details needed to assess the impact of specific changes associated with the application may be needed (Applications 4.1.1, 4.2.1, 4.2.3, 4.2.4, 4.2.5, 4.3.1, and 4.3.2).

SY-B16-S1 Time-dependent failure models are useful for the equipment risk significance evaluation and evaluation of the risk impact of changes to QA requirements (Applications 4.3.1 and 4.3.2).

HR-E01-S2 While it is not yet general practice to include errors of commission in the base PSA, it is advantageous to use information on the general causes of errors of commission to reduce the potential for introducing changes that could increase the likelihood of, or create conditions conducive to, errors of commission (all Applications in Category 4).

HR-G01-S1 Performing a detailed assessment of all HEPs may be useful in prioritising the NPP upgrades, back-fitting activities and plant modifications (Application 4.1.1).

DA-B01-S1 A finer level of resolution in identifying groups of components is helpful because too broad grouping can mask the specific features of group members (Application Group 4.2).

DA-D05-S1 A realistic representation of surveillance and maintenance activities is useful for assessing the impact of specific changes (Applications 4.1.1 and Application Group 4.2).

DA-F01-S1 DA-F02-S1

Because technical specifications such as allowed outage times relate to individual trains of systems, a detailed modelling of CCFs reflecting the number of redundancies/diversities and a realistic and specific modelling of CCF events is helpful for assessing the impact of common cause failures (Application Group 4.2).

DF-F01-S1 DF-G01-S1

Analysis of the risk impact of NPP upgrades and justification for lifetime extension may benefit from detailed identification and modelling of possible dependencies and availability of database containing information on different dependencies (Applications 4.1.1 and 4.1.2).

MQ-B03-S2 MQ-D01-S1

Use of reduced truncation values in relation to the particular IEs of major interest may be useful for the assessment of benefits from specific NPP upgrades (Application 4.1.1).

126

Table 13.5 Oversight Activities



IE-A06-S1 IE-B01-S1 IE-C02-S1

The lists of IEs and IE groups should be detailed enough to allow the evaluation of inspection findings and operational events (Applications 5.2.1 and 5.2.2).

SC-B01-S1 The use of proven computer codes and realistic models helps to avoid conservative and simplifying success criteria; this is essential for the assessment of a particular inspection finding or event (Application Group 5.2).


IE-F04-S1 The time trend analysis in relation to IEs provides useful information for decision making (Application Group 5.1).

IE-H02-S1 An electronic IE database is useful for Application Group 5.1.

All AS-SAs Detailed realistic representation of plant behaviour addressing different mitigation strategies would be helpful for adequate reflection of a wider range of inspection findings and operational events in the PSA model (Applications 5.2.1 and 5.2.2).

SY-B13-S1 SY-B15-S1 SY-C15-S1

More components may need to be included into the model for Application Group 5.2.

DA-B01-S1 DA-F01-S1 DA-F02-S1 DA-H02-S1

These SAs would result in a more detailed PSA and better understanding of particular performance issues (Applications 5.1.2, 5.1.3, Application Group 5.2).

DA-E01-S1 The time trend analysis is useful for the assessment of long term performance indicators (Application 5.1.2).

DF-F01-S1 DF-G01-S1

Detailed identification and modelling of possible dependencies and availability of database containing information on different dependencies is helpful for all Applications in Category 5.

MQ-B03-S1 MQ-B03-S2 MQ-D01-S1

Application of reduced truncation values may be useful for planning the inspection activities (Application 5.1.1), for the analysis of performance indicators (Applications 5.1.2 and 5.1.3), and for possibility to assess inspection findings and operational events (Applications 5.2.1 and 5.2.2).

MQ-C02-S1 A more detailed modelling of specific system configurations may be useful for planning the inspection activities and assessment of risk-based performance indicators (Application Group 5.1).

127

Table 13.6 Evaluation of Safety Issues



All SAs For Application Category 6 any of the special attributes may be required on a case-by-case basis depending on the issue to be analyzed. For these applications, the PSA model should be upgraded in a manner that would allow evaluating the impact associated with the considered measure or issue by the PSA model.


None

128

T

able

13.

7 M

appi

ng th

e Sp

ecia

l Attr

ibut

es o

f PSA

Ele

men

ts to

PSA

App

licat

ions

8

PSA

Ele

men

ts

PSA

App

licat

ion

Cat

egor

y PS

A A

pplic

atio

n G

roup

/ PS

A A

pplic

atio

n IE

A

S SC

SY

H

R

DA

D

F M

Q

RI

1.1

Ass

essm

ent o

f the

ove

rall

plan

t saf

ety

- -

- -

- -

- -

-

1.2

Per

iodi

c sa

fety

revi

ew

IE-F

04-S

1 IE

-H02

-S1

AS-

C02

-S1

AS-

C12

-S1

AS-

C15

-S1

SC-B

01-S

1 -

HR

-G04

-S1

DA

-E01

-S1

DA

-H02

-S1

DF-

F01-

S1

DF-

G01

-S1

- -

1. S

AFE

TY

A

SSE

SSM

EN

T

1.3

Ana

lysi

s of t

he d

egre

e of

de

fenc

e ag

ains

t ass

umed

te

rror

ist a

ttack

scen

ario

s

IE-A

06-S

1 IE

-B01

-S1

AS-

C04

-S1

AS-

C15

-S1

- -

- -

- -

-

2.1

App

licat

ion

of P

SA to

su

ppor

t dec

isio

ns m

ade

durin

g th

e N

PP d

esig

n

(pla

nt u

nder

des

ign)

IE-A

06-S

1 IE

-C02

-S1

IE-C

02-S

2

AS-

B03

-S1

AS-

C02

-S1

AS-

C03

-S1

AS-

C12

-S1

AS-

C14

-S1

AS-

C15

-S1

- SY

-C07

-S1

SY-C

14-S

1 SY

-C15

-S1

SY-B

13-S

1 SY

-B15

-S1

HR

-E01

-S2

DA

-F01

-S1

DA

-C02

-S1

D

F-F0

1-S1

-

- 2.

DE

SIG

N

EV

AL

UA

TIO

N

2.2

Ass

essm

ent o

f the

safe

ty

impo

rtanc

e of

dev

iatio

ns

betw

een

an e

xist

ing

plan

t de

sign

and

upd

ated

/revi

sed

dete

rmin

istic

des

ign

rule

s

- -

SC-B

01-S

1 SY

-C07

-S1

SY-C

14-S

1 SY

-C15

-S1

SY-B

13-S

1 SY

-B15

-S1

- D

A-F

01-S

1 D

F-F0

1-S1

D

F-G

01-S

1 M

Q-B

03-S

1 M

Q-B

03-S

2 M

Q-D

01-S

1

-

3.1

NPP

mai

nten

ance

3.1.

1 M

aint

enan

ce p

rogr

am

optim

izat

ion

IE

-A06

-S1

IE-B

01-S

1 -

- SY

-B08

-S1

SY-B

08-S

2 SY

-B13

-S1

SY-B

15-S

1

HR

-D06

-S1

DA

-B01

-S1

DA

-D05

-S1

DF-

F01-

S1

DF-

G01

-S1

MQ

-C02

-S1

-

3.1.

2 R

isk-

info

rmed

hou

se

keep

ing

- -

- -

-

- -

-

3. N

PP O

PER

AT

ION

3.1.

3 R

isk-

info

rmed

supp

ort f

or

plan

t age

ing

man

agem

ent

prog

ram

IE-F

04-S

1 IE

-H02

-S1

- -

SY-B

13-S

1 SY

-B15

-S1

- D

A-E

01-S

1 D

F-F0

1-S1

D

F-G

01-S

1 -

-

8 The

gen

eral

attr

ibut

es d

escr

ibed

in S

ectio

ns 4

-12

of th

e pu

blic

atio

n ar

e co

nsid

ered

app

licab

le to

all

PSA

app

licat

ions

and

are

not

refe

rred

in th

e ta

ble.

Onl

y th

e sp

ecia

l attr

ibut

es to

be

met

in a

dditi

on to

th

e ge

nera

l at

tribu

tes

are

depi

cted

. Th

e id

entif

iers

of

the

attri

bute

s th

at a

re c

onsi

dere

d es

sent

ial

for

the

appl

icat

ions

(i.e

. es

sent

ial

spec

ial

attri

bute

s in

trodu

ced

in S

ectio

n 13

), ar

e pr

ovid

ed i

n bo

ld

unde

rlin

ed f

ont.

Thos

e at

tribu

tes,

whi

ch c

ould

be

help

ful f

or th

e ap

plic

atio

ns, b

ut a

re n

ot d

eem

ed v

ery

impo

rtant

in a

ccor

danc

e w

ith th

e cu

rren

t sta

te o

f th

e ar

t (i.

e. s

uppl

emen

tal s

peci

al a

ttrib

utes

in

trodu

ced

in S

ectio

n 13

), ar

e pr

ovid

ed in

regu

lar f

ont.

129

PSA

Ele

men

ts

PSA

App

licat

ion

Cat

egor

y PS

A A

pplic

atio

n G

roup

/ PS

A A

pplic

atio

n IE

A

S SC

SY

H

R

DA

D

F M

Q

RI

3.2

Acc

iden

t miti

gatio

n an

d em

erge

ncy

plan

ning

3.2.

1 D

evel

opm

ent a

nd

impr

ovem

ent o

f the

em

erge

ncy

oper

atin

g pr

oced

ures

- A

S-B

03-S

1 A

S-C

02-S

1 A

S-C

03-S

1 A

S-C

14-S

1 A

S-C

04-S

1 A

S-C

05-S

1 A

S-C

06-S

1

- -

HR

-G02

-S1

HR

-G04

-S1

HR

-E01

-S1

- D

F-F0

1-S1

D

F-G

01-S

1 -

-

3.2.

2 S

uppo

rt fo

r NPP

acc

iden

t m

anag

emen

t (se

vere

ac

cide

nt p

reve

ntio

n,

seve

re a

ccid

ent

miti

gatio

n)

IE-A

06-S

1 IE

-B01

-S1

- -

- H

R-G

02-S

1 -

DF-

F01-

S1

DF-

G01

-S1

- -

3.2.

3 S

uppo

rt fo

r NPP

em

erge

ncy

plan

ning

IE

-A06

-S1

IE-B

01-S

1 -

- -

- -

- -

-

3.

3 P

erso

nnel

trai

ning

3.

3.1

Impr

ovem

ent o

f ope

rato

r tra

inin

g pr

ogra

m

IE-A

06-S

1 IE

-F04

-S1

AS-

B03

-S1

AS-

C02

-S1

AS-

C03

-S1

AS-

C14

-S1

AS-

C04

-S1

AS-

C05

-S1

AS-

C06

-S1

AS-

C15

-S1

- SY

-B08

-S1

SY-B

08-S

2 SY

-B13

-S1

SY-B

15-S

1

HR

-G04

-S1

HR

-E01

-S2

DA

-D05

-S1

- M

Q-C

02-S

1 -

3.

3.2

Impr

ovem

ent o

f m

aint

enan

ce p

erso

nnel

tra

inin

g pr

ogra

m

IE-A

06-S

1 IE

-F04

-S1

AS-

C15

-S1

- SY

-B08

-S1

HR

-D06

-S1

- -

- -

3.

3.3

Impr

ovem

ent o

f pla

nt

man

agem

ent t

rain

ing

prog

ram

IE-A

06-S

1 IE

-F04

-S1

AS-

C02

-S1

AS-

C03

-S1

AS-

C14

-S1

AS-

C04

-S1

AS-

C05

-S1

AS-

C06

-S1

AS-

C15

-S1

- -

HR

-G04

-S1

DA

-D05

-S1

- M

Q-C

02-S

1 -

130

PSA

Ele

men

ts

PSA

App

licat

ion

Cat

egor

y PS

A A

pplic

atio

n G

roup

/ PS

A A

pplic

atio

n IE

A

S SC

SY

H

R

DA

D

F M

Q

RI

3.

4 R

isk-

base

d co

nfig

urat

ion

cont

rol/

Ris

k M

onito

rs

3.

4.1

Con

figur

atio

n pl

anni

ng

(e.g

. sup

port

for p

lant

m

aint

enan

ce a

nd te

st

activ

ities

)

IE-A

06-S

1 IE

-F04

-S1

- -

SY-B

08-S

1 SY

-B08

-S2

SY-B

08-S

3 SY

-B16

-S1

- D

A-B

01-S

1 D

A-D

05-S

1 D

A-F

01-S

1 D

A-F

02-S

1

DF-

F01-

S1

DF-

G01

-S1

MQ

-C02

-S1

-

3.

4.2

Rea

l tim

e co

nfig

urat

ion

asse

ssm

ent a

nd c

ontro

l (r

espo

nse

to e

mer

ging

co

nditi

ons)

- A

S-B

03-S

1 A

S-C

02-S

1 A

S-C

03-S

1 A

S-C

14-S

1 A

S-C

04-S

1 A

S-C

05-S

1 A

S-C

06-S

1 A

S-C

15-S

1

- SY

-B08

-S1

SY-B

08-S

2 SY

-B08

-S3

- D

A-B

01-S

1 D

A-D

05-S

1 D

A-F

01-S

1 D

A-F

02-S

1

DF-

F01-

S1

DF-

G01

-S1

MQ

-C02

-S1

-

3.

4.3

Exe

mpt

ions

to T

S an

d ju

stifi

catio

n fo

r con

tinue

d op

erat

ion

IE-B

01-S

1 -

- SY

-B08

-S1

SY-B

08-S

2 SY

-B08

-S3

- D

A-B

01-S

1 D

A-D

05-S

1 D

A-F

01-S

1 D

A-F

02-S

1

DF-

F01-

S1

DF-

G01

-S1

MQ

-C02

-S1

-

3.

4.4

Dyn

amic

risk

-info

rmed

TS

-

- -

SY-B

08-S

1 SY

-B08

-S2

SY-B

08-S

3

- D

A-B

01-S

1 D

A-D

05-S

1 D

A-F

01-S

1 D

A-F

02-S

1

DF-

F01-

S1

DF-

G01

-S1

MQ

-C02

-S1

-

4.1

Pla

nt c

hang

es

4.1.

1 N

PP u

pgra

des,

back

-fit

ting

activ

ities

and

pla

nt

mod

ifica

tions

IE-F

04-S

1 IE

-H02

-S1

IE-A

06-S

1 IE

-B01

-S1

IE-C

02-S

1 IE

-C02

-S2

IE-H

02-S

1

AS-

B03

-S1

AS-

C02

-S1

AS-

C03

-S1

AS-

C14

-S1

AS-

C04

-S1

AS-

C05

-S1

AS-

C06

-S1

AS-

C15

-S1

SC-B

01-S

1 SY

-B16

-S1

SY-B

13-S

1 SY

-B15

-S1

SY-C

15-S

1

HR

-E01

-S2

HR

-G01

-S1

DA

-C02

-S1

DA

-D05

-S1

DF-

F01-

S1

DF-

G01

-S1

MQ

-B03

-S2

M

Q-D

01-S

1 -

4. P

ER

MA

NE

NT

C

HA

NG

ES

TO

TH

E

OPE

RA

TIN

G

PLA

NT

4.1.

2 L

ifetim

e ex

tens

ion

IE-F

04-S

1 IE

-H02

-S1

IE-B

01-S

1 IE

-F04

-S1

IE-H

02-S

1

- -

- H

R-E

01-S

2 D

A-E

01-S

1 D

F-F0

1-S1

D

F-G

01-S

1 -

-

131

PSA

Ele

men

ts

PSA

App

licat

ion

Cat

egor

y PS

A A

pplic

atio

n G

roup

/ PS

A A

pplic

atio

n IE

A

S SC

SY

H

R

DA

D

F M

Q

RI

4.

2 T

echn

ical

spec

ifica

tion

chan

ges

4.

2.1

Det

erm

inat

ion

and

eval

uatio

n of

cha

nges

to

allo

wed

out

age

time

and

chan

ges t

o re

quire

d TS

ac

tions

IE-A

06-S

1 -

- SY

-B08

-S1

SY-B

08-S

2 SY

-B08

-S3

SY-B

16-S

1 SY

-B13

-S1

SY-B

15-S

1 SY

-C15

-S1

HR

-E01

-S2

DA

-B01

-S1

DA

-D05

-S1

DA

-F01

-S1

DA

-F02

-S1

- M

Q-C

02-S

1 -

4.

2.2

Ris

k-in

form

ed

optim

isat

ion

of T

S IE

-A06

-S1

- -

SY-B

08-S

3 SY

-B16

-S1

SY-B

13-S

1

HR

-E01

-S2

DA

-B01

-S1

DA

-D05

-S1

DA

-F01

-S1

DA

-F02

-S1

- M

Q-C

02-S

1 -

4.

2.3

Det

erm

inat

ion

and

eval

uatio

n of

cha

nges

to

surv

eilla

nce

test

inte

rval

s

IE-A

06-S

1 -

- SY

-B08

-S3

SY-B

16-S

1 SY

-B13

-S1

SY-B

15-S

1 SY

-C15

-S1

HR

-E01

-S2

DA

-B01

-S1

DA

-D05

-S1

DA

-F01

-S1

DA

-F02

-S1

- M

Q-C

02-S

1 -

4.

2.4

Ris

k-in

form

ed in

-ser

vice

te

stin

g (I

ST)

IE-A

06-S

1 IE

-C02

-S1

IE-C

02-S

2

- -

SY-B

16-S

1 SY

-B08

-S1

SY-B

08-S

2 SY

-B15

-S1

SY-C

15-S

1

HR

-E01

-S2

DA

-B01

-S1

DA

-D05

-S1

DA

-F01

-S1

DA

-F02

-S1

- M

Q-C

02-S

1 -

4.

2.5

Ris

k-in

form

ed in

-ser

vice

in

spec

tions

(ISI

) IE

-C02

-S1

IE-C

02-S

2 -

- SY

-C01

-S1

SY-B

13-S

1 SY

-B15

-S1

SY-C

15-S

1 SY

-B06

-S1

HR

-E01

-S2

DA

-B01

-S1

DA

-D05

-S1

DA

-F01

-S1

DA

-F02

-S1

- M

Q-C

02-S

1 -

132

PSA

Ele

men

ts

PSA

App

licat

ion

Cat

egor

y PS

A A

pplic

atio

n G

roup

/ PS

A A

pplic

atio

n IE

A

S SC

SY

H

R

DA

D

F M

Q

RI

4.

3 E

stab

lishm

ent o

f gra

ded

QA

pro

gram

for

SSC

4.

3.1

Equi

pmen

t ris

k si

gnifi

canc

e ev

alua

tion

IE

-F04

-S1

IE-H

02-S

1 -

- SY

-B08

-S3

SY-B

13-S

1 SY

-B15

-S1

SY-C

15-S

1 SY

-B16

-S1

HR

-E01

-S2

- -

- -

4.

3.2

Eva

luat

ion

of ri

sk im

pact

of

cha

nges

to Q

A

requ

irem

ents

IE-F

04-S

1 IE

-H02

-S1

- -

SY-B

08-S

3 SY

-B13

-S1

SY-B

15-S

1 SY

-C15

-S1

SY-B

16-S

1

HR

-E01

-S2

- -

- -

5.1

Per

form

ance

mon

itori

ng

5.1.

1 P

lann

ing

and

prio

ritiz

atio

n of

insp

ectio

n ac

tiviti

es

(reg

ulat

ory

and

indu

stry

)

IE-F

04-S

1 IE

-H02

-S1

- -

- -

- D

F-F0

1-S1

D

F-G

01-S

1 M

Q-B

03-S

1 M

Q-B

03-S

2 M

Q-C

02-S

1 M

Q-D

01-S

1

-

5.1.

2 L

ong

term

risk

-bas

ed

perfo

rman

ce in

dica

tors

IE

-F04

-S1

IE-H

02-S

1 -

- -

- D

A-B

01-S

1 D

A-F

01-S

1 D

A-F

02-S

1 D

A-H

02-S

1 D

A-E

01-S

1

DF-

F01-

S1

DF-

G01

-S1

MQ

-B03

-S1

M

Q-B

03-S

2

MQ

-C02

-S1

M

Q-D

01-S

1

-

5. O

VE

RSI

GH

T

AC

TIV

ITIE

S

5.1.

3 S

hort

term

risk

-bas

ed

perfo

rman

ce in

dica

tors

IE

-F04

-S1

IE-H

02-S

1 -

- -

- D

A-B

01-S

1 D

A-F

01-S

1 D

A-F

02-S

1 D

A-H

02-S

1

DF-

F01-

S1

DF-

G01

-S1

MQ

-B03

-S1

MQ

-B03

-S2

MQ

-C02

-S1

MQ

-D01

-S1

-

5.

2 P

erfo

rman

ce a

sses

smen

t

5.

2.1

Ass

essm

ent o

f ins

pect

ion

findi

ngs

IE-A

06-S

1 IE

-B01

-S1

IE-C

02-S

1

All

SAs f

or

AS

SC-B

01-S

1 SY

-B13

-S1

SY-B

15-S

1 SY

-C15

-S1

- D

A-B

01-S

1 D

A-F

01-S

1 D

A-F

02-S

1 D

A-H

02-S

1

DF-

F01-

S1

DF-

G01

-S1

MQ

-B03

-S1

M

Q-B

03-S

2

MQ

-D01

-S1

-

5.

2.2

Eva

luat

ion

and

ratin

g of

op

erat

iona

l eve

nts

IE-A

06-S

1 IE

-B01

-S1

IE-C

02-S

1

All

SAs f

or

AS

SC-B

01-S

1 SY

-B13

-S1

SY-B

15-S

1 SY

-C15

-S1

- D

A-B

01-S

1 D

A-F

01-S

1 D

A-F

02-S

1 D

A-H

02-S

1

DF-

F01-

S1

DF-

G01

-S1

MQ

-B03

-S1

M

Q-B

03-S

2

MQ

-D01

-S1

-

133

PSA

Ele

men

ts

PSA

App

licat

ion

Cat

egor

y PS

A A

pplic

atio

n G

roup

/ PS

A A

pplic

atio

n IE

A

S SC

SY

H

R

DA

D

F M

Q

RI

6.1

Ris

k ev

alua

tion

6.1.

1 R

isk

eval

uatio

n of

co

rrec

tive

mea

sure

s A

ny o

f the

spec

ial a

ttri

bute

s may

be

requ

ired

on

a ca

se-b

y-ca

se b

asis

. T

he P

SA m

odel

shou

ld b

e up

grad

ed in

the

man

ner

that

wou

ld a

llow

an

eval

uatio

n of

the

impa

ct a

ssoc

iate

d w

ith th

e co

nsid

ered

mea

sure

by

the

PSA

mod

el.

-

6. E

VA

LU

AT

ION

O

F SA

FET

Y IS

SUE

S

6.1.

2 R

isk

eval

uatio

n to

iden

tify

and

rank

safe

ty is

sues

A

ny o

f the

spec

ial a

ttri

bute

s may

be

requ

ired

on

a ca

se-b

y-ca

se b

asis

. T

he P

SA m

odel

shou

ld b

e up

grad

ed in

the

man

ner

that

wou

ld a

llow

an

eval

uatio

n of

the

impa

ct a

ssoc

iate

d w

ith th

e co

nsid

ered

issu

e by

the

PSA

mod

el.

-

6.

2 R

egul

ator

y de

cisi

ons

6.

2.1

Lon

g te

rm re

gula

tory

de

cisi

ons

Any

of t

he sp

ecia

l att

ribu

tes m

ay b

e re

quir

ed o

n a

case

-by-

case

bas

is.

The

PSA

mod

el sh

ould

be

upgr

aded

in th

e m

anne

r th

at w

ould

allo

w a

n ev

alua

tion

of th

e im

pact

ass

ocia

ted

with

the

cons

ider

ed m

easu

re b

y th

e PS

A m

odel

. -

6.

2.2

Inte

rim re

gula

tory

de

cisi

ons

Any

of t

he sp

ecia

l att

ribu

tes m

ay b

e re

quir

ed o

n a

case

-by-

case

bas

is.

The

PSA

mod

el sh

ould

be

upgr

aded

in th

e m

anne

r th

at w

ould

allo

w a

n ev

alua

tion

of th

e im

pact

ass

ocia

ted

with

the

cons

ider

ed is

sue

by th

e PS

A m

odel

. -

134

14. CONCLUSIONS

The matter of PSA quality is very important from the viewpoint of risk-informed decision making on various aspects of NPP operation and licensing. This publication provides an approach for achieving the technical consistency of PSA in order to support reliably various PSA applications. The approach involves the consideration of a set of technical features, called attributes, of the major PSA elements relevant for various applications.

A comprehensive list of PSA applications has been compiled. The applications were

grouped into the six categories. Some of the categories include several groups. For each PSA application, a brief description of the purpose of the application and the way the PSA can be used to support it were provided along with the information on what PSA results and metrics can be used in the decision making process.

This publication covers a Level-1 internal events at-power PSA. Nine PSA elements

characterizing the major PSA tasks were defined. For each PSA element, a set of general attributes needed for all PSA applications and special attributes needed for specific PSA applications were elaborated. In relation to each PSA application, the special attributes were further distinguished as essential and supplemental. An essential special attribute emphasises a feature of the PSA element that is considered important to generate the results needed to reliably support the PSA application. A supplemental special attribute may not be strongly required for a specific application, but could further enhance the usefulness of the PSA by providing a greater level of detail, or improving confidence in the results. The publication provides a mapping of the special attributes to the considered PSA applications with a brief explanation on why a special attribute is needed in each particular case.

This publication can be used by PSA practitioners for appropriate planning of PSA

projects taking into account possible uses of the PSA in the future. The publication can be also used by reviewers as an aid in assessing the quality of PSAs and judging the adequacy of a PSA for specific applications. Particularly, the publication can be used by regulatory authorities to support regulatory reviews of licensees’ PSA and PSA application cases along with other IAEA publications, e.g. Refs [19], [20].

135

Appendix I

RISK METRIC DEFINITIONS

Importance measures - Risk achievement worth (RAW)

This is the ratio or interval of the figure of merit, evaluated with the SSC’s basic event probability set to one, to the base case figure of merit. This importance measure reflects the increase in a selected figure of merit when an SSC is assumed to be unable to perform its function due to testing, maintenance, or failure.

- Fussell-vesely (FV)

For a specific basic event, the FV importance is the fractional contribution to the total of a selected figure of merit for all accident sequences containing the basic event to be evaluated.

Core damage frequency, annual average (CDFAVE) This is the frequency of core damage that is averaged over the time-dependent variations that may be exhibited during the course of a year due to plant configuration changes, removing equipment from service to perform tests or maintenance, and the occurrence of plant initiating events which may in fact vary over the course of a reactor lifetime. Periodic updates of this risk metric over the course of the plant lifetime provide a slow version of a time dependent Risk Monitor that reflects broad trends in plant and SSC performance that are reflected in the plant data as well as any permanent changes that are made in the design, maintenance and operation. Change in core damage frequency, annual average (ΔCDFAVE) This is the change in the annual average CDF due some change that is being evaluated that may impact this risk metric. The change may be due to an observed degradation, design change, procedure change, change in test, maintenance or inspection practice, change in performance of an SSC, or changes to any input or assumption associated with the PSA model. Core damage frequency, time-dependent (CDF{t}) This is the frequency of core damage as a function of time. It is often referred to as the ‘instantaneous core damage frequency.’ Only some of the parameters that the CDF is dependent on can be monitored in a time-dependent fashion, such as the time periods when equipment is removed from service. Core damage probability (CDP) This is the total probability of a core damage event over a specified time interval.

137

Conditional core samage probability (CCDP) This is the conditional core damage probability given the occurrence of an initiating event. It is calculated by selecting the appropriate PSA initiating event, setting the initiating event frequency to 1, and solving the PSA model for core damage for the condition that the initiating event has occurred. Incremental conditional core damage probability (ICCDP) This is the increase in the CDP, over that expected from the Base CDF during a configuration change (denoted by the index j) within the time Tj with an increased CDFj relative to CDFBASE. It is referred to as a conditional probability because it is conditioned on being in a specific plant configuration. Large early release frequency, annual average (LERFAVE) This is the frequency of a large early release that is averaged over the time-dependent variations that may be exhibited during the course of a year, which may in fact vary over the course of a reactor lifetime. Change in large early release frequency, annual average (ΔLERFAVE) This is the change in the annual average LERF due some change that is being evaluated that may impact this risk metric. The change may be due to an observed degradation, design change, procedure change, change in test, maintenance or inspection practice, change in performance of an SSC, or changes to any input or assumption associated with the PSA model. Large early release frequency, time-dependent (LERF{t}) This is the frequency of a large early release as a function of time. It is often referred to as the ‘instantaneous large early release frequency.’ Only some of the parameters that the LERF is dependent on can be monitored in a time-dependent fashion, such as the time periods when equipment is removed from service. Large early release probability (LERP) This is the total probability of a large early release over a specified time interval. Conditional large early release probability (CLERP) This is the conditional large early release probability given the occurrence of an initiating event. It is calculated by selecting the appropriate PSA initiating event, setting the initiating event frequency to 1, and solving the PSA model for large early release under the condition that the initiating event has occurred. Incremental conditional large early release probability (ICLERP) This is the increase in the LERP, over that expected from the Base LERF during a configuration change (denoted by the index j) within the time Tj with an increased LERFj

138

relative to LERFBASE. It is referred to as a conditional probability because it is conditioned on being in a specific plant configuration. Quantitative health objectives (QHO) QHOs are derived from qualitative safety goals and provide a numerical measure of acceptable levels of risk of acute and latent health effects due to accidents. These are normally expressed in terms of a small fraction, e.g. .01% of the expected annual risks of death to individuals in designated areas around the plant due to other causes. Accidental risks of death due to radiation sickness are normally compared against non-nuclear accident risks to an average individual within short distances of the plant (e.g. 1 mile), whereas risks due to latent health effects from a reactor accident are normally measured against the risk of latent cancer fatalities due to non-nuclear causes to the population over larger distances from the plant (e.g. 10 miles). Individual risks from NPP accidents are computed by summing the products of the accident frequencies and total estimate of consequences in the appropriate population segment and then normalizing this population risk metric by the population in that segment.

139

Appendix II

PSA APPLICATIONS

143

T

able

A-I

I D

escr

iptio

n of

PSA

App

licat

ions

Brie

f des

crip

tion

of P

SA a

pplic

atio

n PS

A re

sults

and

m

etric

s for

use

in

deci

sion

mak

ing9

Com

men

ts o

n ho

w P

SA m

odel

s can

be

used

to su

ppor

t app

licat

ion

and

exam

ples

1. S

AFE

TY

ASS

ESS

ME

NT

1.1

Ass

essm

ent o

f the

ove

rall

plan

t saf

ety

The

asse

ssm

ent o

f the

ove

rall

plan

t saf

ety

repr

esen

ts th

e m

ain

purp

ose

of P

SA p

erfo

rman

ce a

nd in

clud

es

iden

tific

atio

n an

d ra

nkin

g of

impo

rtant

des

ign

and

oper

atio

nal f

eatu

res,

of d

omin

ant a

ccid

ent s

eque

nces

, sy

stem

s, co

mpo

nent

s, hu

man

inte

ract

ions

and

de

pend

enci

es im

porta

nt fo

r saf

ety.

A c

ompa

rison

of t

he

resu

lts a

gain

st sa

fety

goa

ls o

r qua

ntita

tive

heal

th o

bjec

tives

m

ay b

e in

volv

ed.

CD

F AV

E, LE

RF A

VE,

QH

Os,

risk

impo

rtanc

e m

easu

res o

f all

SSC

s and

HEs

, pr

imar

y co

ntrib

utor

s to

risk

A c

ompl

ete

safe

ty e

valu

atio

n w

ould

in p

rinci

ple

requ

ire a

full

scop

e PS

A (L

evel

-1 th

roug

h 3)

co

verin

g al

l ope

ratin

g an

d sh

utdo

wn

mod

es a

nd b

oth

inte

rnal

and

ext

erna

l pla

nt h

azar

ds.

A

parti

al b

ut m

eani

ngfu

l eva

luat

ion

can

be p

erfo

rmed

usi

ng a

lim

ited

scop

e PS

A w

ith q

ualit

ativ

e ev

alua

tion

of th

e m

issi

ng sc

ope

supp

orte

d by

bou

ndin

g as

sess

men

ts. R

isk

cont

ribut

ors a

nd

impo

rtanc

e in

form

atio

n ar

e us

ed to

dev

elop

risk

insi

ghts

.

1.2

Per

iodi

c sa

fety

rev

iew

Si

mila

r to

Item

1.1

, PSA

pro

vide

s use

ful i

nsig

hts t

o su

ppor

t a p

erio

dic

safe

ty re

view

. A sa

fety

ass

essm

ent

proc

ess c

onsi

sts i

n id

entif

ying

safe

ty is

sues

, det

erm

inin

g th

eir s

afet

y si

gnifi

canc

e an

d m

akin

g de

cisi

ons o

n th

e ne

ed

for c

orre

ctiv

e m

easu

res.

A m

ajor

ben

efit

of in

clud

ing

PSA

in

per

iodi

c re

view

s is t

he c

reat

ion

of a

n up

-to-d

ate

over

view

of t

he w

hole

pla

nt. P

SA m

ay h

elp

in

iden

tific

atio

n of

real

cos

t-eff

ectiv

e im

prov

emen

ts to

safe

ty.

The

appl

icat

ion

may

invo

lve

mod

ellin

g of

agi

ng e

ffec

ts in

PS

A. T

ypic

ally

ther

e ar

e tw

o ty

pes o

f SSC

s: (1

) SSC

s w

hich

are

per

iodi

cally

rene

wed

or r

epla

ced

(can

be

mod

elle

d ‘a

s new

’), a

nd (2

) SSC

s whi

ch a

re su

bjec

t to

an

agin

g pr

oces

s. C

urre

ntly

, mod

ellin

g of

SSC

agi

ng in

term

s of

PSA

is in

an

expl

orat

ory

stag

e. T

hus,

PSA

trea

tmen

t of

agin

g ef

fect

s is n

ot st

anda

rdly

car

ried

out a

nd is

con

side

red

beyo

nd th

e cu

rren

t sta

te o

f the

art.

In so

me

coun

ties,

PSA

is

requ

ired

for l

ifetim

e ex

tens

ion

and

to su

ppor

t a c

ost

bene

fit e

valu

atio

n of

pos

sibl

e ba

ckfit

s to

redu

ce th

e ris

k of

se

vere

acc

iden

ts.

Agi

ng e

ffec

ts, h

owev

er, a

re ty

pica

lly

addr

esse

d qu

alita

tivel

y.

ΔC

DF A

VE,

ΔLE

RF A

VE,

CD

F AV

E, LE

RF A

VE,

QH

Os,

risk

impo

rtanc

e m

easu

res o

f all

SSC

s and

HEs

, pr

imar

y co

ntrib

utor

s to

risk

In a

dditi

on to

the

disc

ussi

on a

bove

in It

em 1

.1, t

he im

porta

nt is

sues

in th

is a

pplic

atio

n ar

e th

e us

e of

pla

nt-s

peci

fic d

ata,

mod

ellin

g of

as-

built

-as-

oper

ated

pla

nt c

ondi

tions

, and

add

ress

ing

the

poss

ible

impa

ct o

f agi

ng p

heno

men

a an

d co

mpo

nent

life

time

cons

ider

atio

ns o

n th

e ov

eral

l ris

k m

etric

s. Se

nsiti

vity

cal

cula

tions

may

be

requ

ired

to a

sses

s the

pot

entia

l effe

ct o

f age

ing

on

pass

ive

com

pone

nts,

whi

ch a

re n

ot n

orm

ally

mai

ntai

ned

or re

plac

ed.

For t

he c

ost b

enef

it ev

alua

tion

of se

vere

acc

iden

t man

agem

ent a

ltern

ativ

es th

e PS

A n

eeds

to b

e up

date

d to

refle

ct d

esig

n an

d pr

oced

ure

chan

ges a

ssoc

iate

d w

ith e

ach

alte

rnat

ive

(Lev

el-2

and

-3

PSA

. The

cha

nges

in ri

sk m

etric

s are

use

d to

eva

luat

e th

e le

vel o

f ris

k re

duct

ion

asso

ciat

ed w

ith

each

alte

rnat

ive

(see

als

o Ite

m 3

.2.2

).

For e

xam

ple,

whi

le li

cens

ing

an N

PP fo

r con

tinua

tion

of it

s ope

ratio

n be

yond

the

desi

gn li

fetim

e,

a fu

ll sc

ope

PSA

is p

erfo

rmed

to e

nsur

e th

at it

mee

ts th

e co

untry

’s Q

HO

s or s

afet

y go

als.

In

addi

tion,

the

prob

lem

s ass

ocia

ted

with

agi

ng m

ay b

e in

vest

igat

ed a

nd a

ccou

nted

for.

In so

me

coun

tries

, it i

s req

uire

d to

com

pare

the

stat

e an

d de

sign

of a

n op

erat

ing

plan

t aga

inst

ac

tual

det

erm

inis

tic re

gula

tions

. The

safe

ty im

porta

nce

of d

evia

tions

can

be

judg

ed b

y th

e he

lp o

f a

plan

t spe

cific

PSA

. The

nee

d of

bac

kfit

mea

sure

s is d

eriv

ed fr

om th

e ou

tcom

e of

this

as

sess

men

t. Se

e al

so It

em 2

.2.

9 Def

initi

ons o

f ris

k m

etric

s are

pro

vide

d in

App

endi

x I.

144

Brie

f des

crip

tion

of P

SA a

pplic

atio

n PS

A re

sults

and

m

etric

s for

use

in

deci

sion

mak

ing9

Com

men

ts o

n ho

w P

SA m

odel

s can

be

used

to su

ppor

t app

licat

ion

and

exam

ples

1.3

Ana

lysi

s of t

he d

egre

e of

def

ence

aga

inst

ass

umed

te

rror

ist a

ttac

k sc

enar

ios

Ass

essm

ent m

ay in

clud

e id

entif

icat

ion

of v

ital p

lant

are

as,

whi

ch m

ay p

ose

sign

ifica

nt ri

sk to

the

plan

t in

case

of

mal

evol

ent a

cts,

and

of e

xplo

rato

ry d

esig

n op

tions

aim

ed

to re

duce

the

risk.

CC

DP,

CLE

RP,

risk

im

porta

nce

mea

sure

s of p

lant

of

all S

SCs,

HEs

, and

pl

ant a

reas

Vita

l are

as id

entif

icat

ion

wou

ld in

prin

cipl

e re

quire

a fu

ll sc

ope

Leve

l 1 th

roug

h 3

PSA

cov

erin

g al

l ope

ratin

g an

d sh

utdo

wn

mod

es a

nd b

oth

inte

rnal

and

ext

erna

l pla

nt h

azar

ds. H

owev

er,

mea

ning

ful r

esul

ts fo

r vita

l are

a id

entif

icat

ion

can

be o

btai

ned

from

Lev

el-1

inte

rnal

and

ext

erna

l ha

zard

s PSA

. A

fter S

epte

mbe

r 11,

200

1, th

is a

pplic

atio

n w

as in

trodu

ced

to a

naly

se th

e qu

antit

ativ

e de

gree

of

defe

nce

of N

PPs a

gain

st a

ttack

scen

ario

s sim

ilar t

o th

e ob

serv

ed. O

ther

type

s of t

erro

rist a

ttack

sc

enar

ios,

i.e. e

xter

nal e

xplo

sion

s can

be

cons

ider

ed w

ithin

the

scop

e of

this

app

licat

ion.

This

app

licat

ion

requ

ires t

he d

evel

opm

ent o

f spe

cial

ana

lysi

s m

odel

s der

ived

from

the

base

line

mod

el b

y ad

ding

eve

nt tr

ees o

r add

ition

al fa

ult t

rees

to th

e ex

istin

g m

odel

incl

udin

g th

e re

quire

d da

ta. I

t req

uire

s a re

asse

ssm

ent o

f the

HR

A fo

r pos

t-acc

iden

t hum

an a

ctio

ns, a

ccid

ent

man

agem

ent a

ctio

ns a

nd S

AM

G-a

ctio

ns, w

ith sp

ecia

l reg

ard

to h

ardw

are

depe

nden

cies

.

Spec

ial d

eter

min

istic

ana

lysi

s in

supp

ort o

f the

mod

el e

xten

sion

s as a

par

t of t

he e

vent

s seq

uenc

e an

alys

is is

requ

ired,

i.e.

for t

he st

ruct

ural

dyn

amic

s of b

uild

ings

, im

pact

of f

ires a

nd e

xplo

sion

s, et

c. E

vent

sequ

ence

ana

lysi

s may

requ

ire a

dditi

onal

exp

erim

enta

l ana

lysi

s.

2. D

ESI

GN

EV

AL

UA

TIO

N

2.1

App

licat

ion

of P

SA to

supp

ort d

ecis

ions

mad

e du

ring

the

NPP

des

ign

(pla

nt u

nder

des

ign)

The

appl

icat

ion

focu

ses o

n id

entif

icat

ion

of d

esig

n w

eakn

esse

s and

effe

ctiv

e ar

eas f

or im

prov

emen

t in

view

of

plan

t ris

k. A

sses

smen

t may

incl

ude

inve

stig

atio

n of

va

riant

s and

exp

lora

tory

des

ign

optio

ns, s

uffic

ienc

y in

sy

stem

s’ re

dund

ancy

and

div

ersi

ty, e

ffect

iven

ess i

n em

erge

ncy

and

acci

dent

man

agem

ent m

easu

res,

as w

ell a

s de

velo

pmen

t of r

elia

bilit

y an

d av

aila

bilit

y ta

rget

s for

SSC

s to

mee

t saf

ety

goal

s, if

set.

CD

F AV

E, LE

RF A

VE,

CD

P, L

ERP,

QH

Os,

ΔC

DF A

VE,

ΔLE

RF A

VE,

Ris

k im

porta

nce

mea

sure

s of a

ffec

ted

SSC

s and

HEs

(e.g

. F-

V, R

AW

), PS

A

insi

ghts

Use

of P

SA m

odel

is si

mila

r to

Item

1.1

exc

ept f

or th

at a

dditi

onal

ass

umpt

ions

are

nee

ded

in li

eu

of la

ck o

f des

ign

and

oper

atio

nal d

etai

ls; u

ncer

tain

ties i

n ris

k es

timat

es a

re c

orre

spon

ding

ly la

rger

th

an fo

r as-

built

pla

nt.

PSA

resu

lts c

an b

e us

ed to

allo

cate

relia

bilit

y ta

rget

s for

SSC

s the

reby

fo

rmin

g pa

rt of

the

desi

gn sp

ecifi

catio

n. F

or d

esig

n ch

ange

eva

luat

ions

, the

leve

l of d

etai

l of t

he

PSA

mod

el in

the

area

s aff

ecte

d by

the

desi

gn c

hang

es m

ay b

e gr

eate

r tha

n th

at fo

r the

rest

of t

he

plan

t.

Parti

cula

r effo

rt sh

ould

be

mad

e to

iden

tify

uniq

ue in

itiat

ing

even

ts, f

ailu

re m

odes

, eve

nt

sequ

ence

s and

dep

ende

ncie

s tha

t may

be

intro

duce

d by

new

des

ign

feat

ures

. Det

aile

d de

pend

ency

m

atric

es sh

ould

be

deve

lope

d to

iden

tify

and

docu

men

t all

phys

ical

and

func

tiona

l dep

ende

ncie

s am

ong

supp

ort s

yste

ms a

nd fr

ont-l

ine

syst

ems.

All

norm

ally

ope

ratin

g sy

stem

s sho

uld

be

exam

ined

to id

entif

y po

ssib

le in

itiat

ing

even

ts th

at m

ay b

e ca

used

by

loss

of t

he e

ntire

syst

em, o

f a

sing

le sy

stem

trai

n, o

r of c

ombi

natio

ns o

f tra

ins.

Even

t seq

uenc

e di

agra

ms m

ay b

e us

ed to

de

velo

p th

e ac

cide

nt se

quen

ces a

nd id

entif

y ch

alle

nges

to re

lief a

nd sa

fety

syst

ems i

n th

e pe

riod

imm

edia

tely

follo

win

g th

e in

itiat

ing

faul

t.

For e

xam

ple,

the

PSA

can

be

used

as a

supp

ortin

g to

ol to

sele

ct o

r mod

ify th

e de

sign

bas

is

acci

dent

s, to

dec

ide

the

clas

sific

atio

n of

safe

ty re

late

d SS

Cs,

to d

efin

e ge

nera

l des

ign

crite

ria, a

nd

to d

evel

op S

SC re

liabi

lity

and

avai

labi

lity

targ

ets.

145

Brie

f des

crip

tion

of P

SA a

pplic

atio

n PS

A re

sults

and

m

etric

s for

use

in

deci

sion

mak

ing9

Com

men

ts o

n ho

w P

SA m

odel

s can

be

used

to su

ppor

t app

licat

ion

and

exam

ples

2.2

Ass

essm

ent o

f the

safe

ty im

port

ance

of d

evia

tions

be

twee

n an

exi

stin

g pl

ant d

esig

n an

d up

date

d/re

vise

d de

term

inis

tic d

esig

n ru

les

Ass

essm

ent m

ay in

clud

e in

vest

igat

ion

of ri

sk-s

igni

fican

ce

of d

evia

tions

from

revi

sed

desi

gn ru

les;

ofte

n pe

rfor

med

in

the

fram

ewor

k of

a p

erio

dic

safe

ty re

view

.

CD

F AV

E, LE

RF A

VE,

QH

Os, Δ

CD

F AV

E, Δ

LER

F AV

E, R

isk

impo

rtanc

e m

easu

res o

f aff

ecte

d SS

Cs a

nd H

Es (e

.g.

F-V

, RA

W)

The

key

issu

e in

this

app

licat

ion

is th

e po

ssib

le im

pact

of c

hang

es in

the

desi

gn ru

les o

n ris

k as

soci

ated

with

pla

nt o

pera

tion.

Spe

cial

mod

el a

djus

tmen

ts, m

odel

ext

ensi

ons a

nd re

visi

ting

key

assu

mpt

ions

may

be

requ

ired

to m

odel

the

devi

atio

n of

the

plan

t des

ign

from

revi

sed

dete

rmin

istic

ru

les o

f con

cern

. Id

entif

icat

ion

of p

rimar

y co

ntrib

utor

s to

risk

and

com

paris

on w

ith sa

fety

targ

et

valu

es (C

DF,

LER

F) m

ay b

e ne

eded

. Fo

r exa

mpl

e, fo

r the

ope

ratin

g pl

ants

con

stru

cted

in a

ccor

danc

e w

ith th

e ol

d de

sign

rule

s, PS

A

resu

lts, a

nd ri

sk m

etric

s can

be

used

to ju

stify

low

risk

sign

ifica

nce

of c

erta

in d

evia

tions

from

re

vise

d ru

les.

3. N

PP O

PER

AT

ION

3.1

NPP

mai

nten

ance

3.1.

1 M

aint

enan

ce p

rogr

am o

ptim

izat

ion

The

appl

icat

ion

incl

udes

ass

essm

ent,

optim

izat

ion

and

esta

blis

hmen

t of m

aint

enan

ce p

lans

and

pro

cedu

res i

n vi

ew

of p

lant

risk

. Mai

nten

ance

act

iviti

es a

re a

sses

sed

to a

ssur

e th

at ri

sk si

gnifi

cant

sys

tem

s and

equ

ipm

ent a

re b

eing

ad

equa

tely

mai

ntai

ned

to su

ppor

t req

uire

d re

liabi

lity,

and

th

at m

aint

enan

ce a

ctiv

ities

do

not r

educ

e pl

ant s

afet

y an

d in

crea

se ri

sk b

y, fo

r exa

mpl

e, e

xten

sive

mai

nten

ance

re

sulti

ng in

incr

ease

d eq

uipm

ent u

nava

ilabi

lity.

Foc

us is

on

equ

ipm

ent w

ith g

reat

est i

mpa

cts o

n pl

ant r

isk.

O

ppor

tuni

ties f

or re

duce

d or

elim

inat

ed m

aint

enan

ce ta

sks

are

defin

ed a

nd e

valu

ated

for S

SCs t

hat h

ave

low

risk

si

gnifi

canc

e an

d fo

r mai

nten

ance

that

doe

s not

supp

ort

criti

cal f

unct

ions

of t

he S

SCs.

Ris

k im

porta

nce

mea

sure

s of a

ffec

ted

SSC

s (e.

g. F

-V,

RA

W), Δ

CD

F AV

E, Δ

LER

F AV

E

Ris

k im

porta

nce

mea

sure

s fro

m th

e ba

se P

SA a

re u

sed

to h

elp

prio

ritiz

e ca

ndid

ate

mai

nten

ance

ch

ange

s: c

hang

es in

CD

F an

d LE

RF

are

used

to ju

stify

acc

epta

ble

risk

impa

cts a

nd to

det

erm

ine

risk

sign

ifica

nce.

Exp

licit

mod

el o

f mai

nten

ance

una

vaila

bilit

ies a

nd c

apab

ility

to p

redi

ct o

r bo

und

the

impa

ct o

f pro

gram

cha

nges

on

failu

re ra

tes a

nd m

aint

enan

ce u

nava

ilabi

litie

s are

nee

ded

to su

ppor

t the

app

licat

ion.

The

leve

l of d

etai

l of t

he P

SA m

odel

in th

e ar

eas a

ffect

ed b

y th

e pr

ogra

m c

hang

es m

ay b

e gr

eate

r tha

n th

at fo

r the

rest

of t

he p

lant

. Fo

r exa

mpl

e, th

e ap

plic

atio

n of

PSA

can

serv

e (a

) to

iden

tify

equi

pmen

t req

uirin

g up

grad

ed

prev

entiv

e m

aint

enan

ce (

as a

n in

crea

se in

its r

elia

bilit

y re

sults

in a

subs

tant

ial g

ain

in sa

fety

), (b

) to

iden

tify

equi

pmen

t req

uirin

g su

stai

ned,

slig

htly

redu

ced

prev

entiv

e m

aint

enan

ce (a

s a d

ecre

ase

in it

s rel

iabi

lity

does

not

affe

ct th

e le

vel o

f saf

ety)

, (c

) to

iden

tify

equi

pmen

t req

uirin

g on

ly

corr

ectiv

e m

aint

enan

ce (a

s its

una

vaila

bilit

y do

es n

ot re

sult

in a

maj

or in

crea

se in

risk

), (d

) to

elim

inat

e m

aint

enan

ce o

n ce

rtain

failu

re m

odes

that

are

not

rele

vant

to th

e ris

k si

gnifi

cant

saf

ety

func

tion,

(e) t

o as

sess

the

impa

ct o

f shi

fting

mai

nten

ance

act

iviti

es fr

om th

e pl

ant o

utag

e to

the

oper

atio

n m

ode

(wou

ld re

quire

a sh

utdo

wn

PSA

).

3.1.

2 R

isk-

info

rmed

hou

se-k

eepi

ng

This

app

licat

ion

is in

tend

ed to

ass

ure

low

risk

co

ntrib

utio

ns fr

om e

xter

nal e

vent

s (e.

g. se

ism

ic) a

nd

inte

rnal

haz

ards

(e.g

. fire

, flo

ods)

by

dire

ctin

g ho

usek

eepi

ng a

ctiv

ities

to ri

sk im

porta

nt a

reas

.

Ris

k im

porta

nce

mea

sure

s of a

ffec

ted

SSC

s (e.

g. F

-V,

RA

W), Δ

CD

F AV

E, Δ

LER

F AV

E

The

resu

lts o

f an

exte

rnal

eve

nt a

nd in

tern

al h

azar

d PS

A a

re u

sed

to a

sses

s the

rela

tive

risk

cont

ribut

ions

of r

oom

s and

are

as a

nd id

entif

y ba

ckfit

ting

mea

sure

s. Fo

r exa

mpl

e, th

e re

sults

of a

n in

tern

al fi

re P

SA m

ay b

e ta

ken

into

acc

ount

for a

djus

ting

trans

ient

co

mbu

stib

le c

ontro

l.

146

Brie

f des

crip

tion

of P

SA a

pplic

atio

n PS

A re

sults

and

m

etric

s for

use

in

deci

sion

mak

ing9

Com

men

ts o

n ho

w P

SA m

odel

s can

be

used

to su

ppor

t app

licat

ion

and

exam

ples

3.1.

3 R

isk-

info

rmed

supp

ort f

or p

lant

age

ing

man

agem

ent p

rogr

am

The

aim

of t

his a

pplic

atio

n is

to o

ptim

ise

the

scop

e of

the

plan

t spe

cific

age

ing

prog

ram

for s

afet

y re

late

d eq

uipm

ent.

It in

clud

es id

entif

icat

ion

of sa

fety

sign

ifica

nt c

ompo

nent

s an

d m

ay in

volv

e m

odel

ling

of a

ging

effe

cts i

n PS

A a

nd

iden

tific

atio

n of

the

risk

sign

ifica

nt S

SCs p

oten

tially

de

grad

ing

due

to a

ging

phe

nom

ena.

Ris

k im

porta

nce

mea

sure

s of a

ll SS

Cs (

e.g.

F-V

, R

AW

), Δ

CD

F AV

E, Δ

LER

F AV

E

The

key

issu

e in

this

app

licat

ion

is th

e po

ssib

le im

pact

of a

ging

phe

nom

ena

and

com

pone

nt

lifet

ime

cons

ider

atio

ns o

n th

e ov

eral

l ris

k m

etric

s. D

epen

ding

on

the

com

pone

nts b

eing

eva

luat

ed

Leve

ls 1

thro

ugh

2 fu

ll-sc

ope

PSA

may

be

need

ed.

3.2

Acc

iden

t miti

gatio

n an

d em

erge

ncy

plan

ning

3.2.

1 D

evel

opm

ent a

nd im

prov

emen

t of t

he e

mer

genc

y op

erat

ing

proc

edur

es

The

syst

emat

ic a

sses

smen

t of p

lant

vul

nera

bilit

ies a

nd th

e in

sigh

ts d

eriv

ed fr

om th

e PS

A p

roce

ss a

re u

sed

to e

stab

lish

or im

prov

e th

e EO

Ps b

y pr

ovid

ing

assu

ranc

e th

at a

bro

ad

scop

e of

vul

nera

bilit

ies i

s add

ress

ed in

a re

alis

tic,

appr

opria

tely

det

aile

d an

d co

nsis

tent

man

ner.

The

inte

gral

vi

ew o

f the

acc

iden

t pro

gres

sion

s pro

vide

s inf

orm

atio

n on

th

e be

nefit

s and

dra

wba

cks o

f var

ious

ope

ratio

ns in

ab

norm

al p

lant

stat

es. T

ypic

ally

, acc

iden

t seq

uenc

e an

alys

is in

PSA

is c

arrie

d ou

t usi

ng e

xist

ing

EOPs

and

as

sess

men

t of a

ssoc

iate

d hu

man

inte

ract

ions

. Thi

s in

turn

pr

ovid

es d

etai

led

info

rmat

ion

for r

econ

side

ring

EOPs

and

ev

entu

al im

prov

emen

ts in

the

light

of P

SA in

sigh

ts. P

SA

can

also

pro

vide

the

basi

s for

spec

ifyin

g th

e de

cisi

on

poin

ts fo

r whe

n th

e tra

nsiti

on in

to th

e SA

MG

pha

se sh

ould

oc

cur.

Ris

k im

porta

nce

mea

sure

s of a

ffec

ted

actio

ns (e

.g. F

-V,

RA

W) a

nd

asso

ciat

ed A

Ss,

ΔC

DF A

VE,

ΔLE

RF A

VE

Impo

rtanc

e m

easu

res f

rom

bas

e PS

A u

sed

to h

elp

prio

ritiz

e ca

ndid

ate

proc

edur

al c

hang

es, c

hang

e in

CD

F an

d LE

RF

used

to ju

stify

acc

epta

ble

risk

impa

cts a

nd to

det

erm

ine

risk

sign

ifica

nce.

The

le

vel o

f det

ail o

f the

PSA

mod

el in

the

area

s aff

ecte

d by

the

proc

edur

e ch

ange

s inc

ludi

ng th

e ac

cide

nt se

quen

ces i

nvok

ing

the

affe

cted

EO

Ps m

ay b

e gr

eate

r tha

n th

at fo

r the

rest

of t

he p

lant

; a

mor

e si

mpl

ified

con

serv

ativ

e tre

atm

ent o

f oth

er p

arts

of t

he m

odel

and

acc

iden

t seq

uenc

es

acce

ptab

le.

The

PSA

mus

t exp

licitl

y re

pres

ent o

pera

tor a

ctio

ns th

at re

fer t

o sp

ecifi

c EO

Ps, a

nd

the

HR

A m

etho

d us

ed in

the

PSA

mus

t be

capa

ble

of p

redi

ctin

g th

e im

pact

of t

he p

roce

dure

ch

ange

s to

supp

ort t

his a

pplic

atio

n.

Exam

ples

incl

ude

deve

lopm

ent a

nd m

odifi

catio

ns o

f eve

nt b

ased

and

sym

ptom

bas

ed p

roce

dure

s, de

velo

ping

pro

cedu

res f

or d

omin

ant a

ccid

ent s

eque

nces

, etc

.

147

Brie

f des

crip

tion

of P

SA a

pplic

atio

n PS

A re

sults

and

m

etric

s for

use

in

deci

sion

mak

ing9

Com

men

ts o

n ho

w P

SA m

odel

s can

be

used

to su

ppor

t app

licat

ion

and

exam

ples

3.2.

2 S

uppo

rt fo

r N

PP a

ccid

ent m

anag

emen

t (se

vere

ac

cide

nt p

reve

ntio

n, se

vere

acc

iden

t miti

gatio

n)

Seve

re a

ccid

ent p

reve

ntio

n: is

bas

ed p

artly

on

PSA

. Ex

istin

g, a

ltern

ativ

e or

add

ition

al sy

stem

s, eq

uipm

ent a

nd

mea

sure

s are

eva

luat

ed a

nd im

plem

ente

d in

the

acci

dent

m

anag

emen

t pro

cedu

res w

ith th

e pu

rpos

e of

rest

orin

g th

e fu

nctio

n of

safe

ty re

late

d sy

stem

s and

for p

reve

ntin

g de

grad

atio

n of

eve

nts i

nto

seve

re a

ccid

ents

.

Seve

re a

ccid

ent m

itiga

tion

incl

udes

PSA

bas

ed

iden

tific

atio

n an

d ca

tego

rizat

ion

of a

ccid

ent s

eque

nces

to

geth

er w

ith d

escr

iptio

ns o

f pla

nt re

spon

ses a

nd

vuln

erab

ilitie

s. PS

A h

elps

to u

nder

stan

d ac

cide

nt

prog

ress

ion,

iden

tific

atio

n of

succ

ess p

aths

and

ass

ocia

ted

stra

tegi

es, p

riorit

isin

g sa

fety

feat

ures

to re

duce

risk

s. Th

e in

tegr

al v

iew

of p

lant

resp

onse

util

ized

in P

SA

met

hodo

logy

is h

elpf

ul in

dis

cern

ing

the

pote

ntia

l for

ne

gativ

e ef

fect

s of c

erta

in m

easu

res.

Ris

k im

porta

nce

mea

sure

s of a

ffec

ted

actio

ns (e

.g. F

-V,

RA

W) a

nd

asso

ciat

ed A

Ss,

ΔC

DF A

VE,

ΔLE

RF A

VE

Impo

rtanc

e m

easu

res f

rom

‘bas

e ca

se P

SA’ a

re u

sed

to h

elp

prio

ritiz

e ca

ndid

ate

acci

dent

m

anag

emen

t pro

cedu

re c

hang

es; c

hang

es in

CD

F an

d LE

RF

are

used

to ju

stify

acc

epta

ble

risk

impa

cts a

nd to

det

erm

ine

risk

sign

ifica

nce.

The

leve

l of d

etai

l of t

he P

SA m

odel

in th

e ar

eas

affe

cted

by

the

chan

ges i

nclu

ding

the

acci

dent

sequ

ence

s inv

okin

g th

e af

fect

ed E

OPs

, AM

Ps, a

nd

affe

cted

SSC

s may

be

grea

ter t

han

that

for t

he re

st o

f the

pla

nt; a

mor

e si

mpl

ified

con

serv

ativ

e tre

atm

ent o

f oth

er p

arts

of t

he m

odel

and

acc

iden

t seq

uenc

es is

acc

epta

ble.

In

the

case

of o

pera

tor

actio

ns to

impl

emen

t acc

iden

t man

agem

ent,

the

PSA

mus

t exp

licitl

y re

pres

ent o

pera

tor a

ctio

ns

that

refe

r to

spec

ific

EOPs

and

AM

Ps, a

nd th

e H

RA

met

hod

used

in th

e PS

A m

ust b

e ca

pabl

e of

pr

edic

ting

the

impa

ct o

f the

pro

cedu

re c

hang

es to

supp

ort t

his a

pplic

atio

n. A

Lev

el 1

PSA

tre

atm

ent o

f ope

rato

r act

ions

can

supp

ort a

ccid

ent m

anag

emen

t pro

cedu

re e

nhan

cem

ent f

or th

ose

actio

ns a

imed

at p

reve

ntin

g se

vere

cor

e da

mag

e, w

hile

a li

mite

d to

full

scop

e Le

vel 2

PSA

is

requ

ired

to a

ddre

ss se

vere

acc

iden

t miti

gatio

n st

rate

gies

. Th

e se

vere

acc

iden

t con

sequ

ence

cod

es

mus

t be

able

to si

mul

ate

the

impl

emen

tatio

n of

the

actio

ns a

nd m

easu

re th

e im

pact

of c

hang

es to

be

eva

luat

ed. I

n th

e ca

se o

f new

har

dwar

e or

des

ign

feat

ures

(e.g

. int

erlo

cks,

new

sign

als,

etc.

) to

impl

emen

t acc

iden

t man

agem

ent r

efer

to It

em 2

.1.

Exam

ples

:

1) S

ever

e ac

cide

nt p

reve

ntio

n: A

typi

cal e

xam

ple

is th

e us

e of

fire

wat

er fo

r coo

ling

safe

ty re

late

d eq

uipm

ent i

f ess

entia

l ser

vice

wat

er is

lost

.

2) S

ever

e ac

cide

nt m

itiga

tion:

A ty

pica

l exa

mpl

e fo

r PW

R re

acto

rs c

onsi

sts i

n re

-fill

ing

of st

eam

ge

nera

tors

with

wat

er d

urin

g a

stea

m g

ener

ator

tube

rupt

ure

initi

ated

seve

re a

ccid

ent i

n or

der t

o en

hanc

e re

tent

ion

of ra

dioa

ctiv

e m

ater

ials

.

3.2.

3 S

uppo

rt fo

r N

PP e

mer

genc

y pl

anni

ng

Bas

ed o

n PS

A, i

mpo

rtant

ele

men

ts o

f em

erge

ncy

plan

ning

ar

e ex

plor

ed a

nd a

ppro

pria

te st

rate

gies

are

dev

elop

ed. T

he

issu

es to

be

expl

ored

are

: the

cha

ract

eris

tics o

f the

pla

nt,

the

plan

t site

and

the

diff

eren

t pos

sibl

e co

unte

rmea

sure

s (s

uch

as sh

elte

ring,

eva

cuat

ion,

iodi

ne p

roph

ylax

is, l

ong

term

relo

catio

n, la

nd d

econ

tam

inat

ion,

food

ban

s). T

his

requ

ires a

Lev

el 3

PSA

. Sp

ecifi

c ap

plic

atio

ns in

clud

e po

ssib

le a

djus

tmen

ts to

the

emer

genc

y pl

anni

ng z

ones

(E

PZ),

refin

emen

t of e

mer

genc

y ac

tion

leve

ls, a

nd

focu

sing

the

reso

urce

s for

eva

cuat

ion

and

shel

terin

g.

QH

Os,

Early

and

la

tent

risk

and

dos

e fo

r diff

eren

t em

erge

ncy

plan

ning

re

spon

ses

This

app

licat

ion

requ

ires a

full

scop

e Le

vel 3

PSA

cov

erin

g al

l mod

es a

nd a

ll in

tern

al a

nd

exte

rnal

pla

nt h

azar

ds a

nd th

e ca

pabi

lity

to p

redi

ct th

e ris

k im

pact

s of a

ny c

hang

es to

the

emer

genc

y pl

an th

at a

re to

be

eval

uate

d in

clud

ing

alte

rnat

ive

evac

uatio

n, sh

elte

ring,

and

food

im

poun

dmen

t stra

tegi

es.

An

exam

ple

is to

supp

ort t

he d

eter

min

atio

n of

EPZ

dis

tanc

e, e

mer

genc

y ac

tion

leve

ls, a

nd

plan

ning

and

trai

ning

for e

vacu

atio

n an

d sh

elte

ring

activ

ities

.

148

Brie

f des

crip

tion

of P

SA a

pplic

atio

n PS

A re

sults

and

m

etric

s for

use

in

deci

sion

mak

ing9

Com

men

ts o

n ho

w P

SA m

odel

s can

be

used

to su

ppor

t app

licat

ion

and

exam

ples

3.3.

Per

sonn

el tr

aini

ng

3.3.

1 Im

prov

emen

t of o

pera

tor

trai

ning

pro

gram

PSA

is u

sed

to im

prov

e op

erat

or tr

aini

ng p

rogr

am b

y pr

ovid

ing

info

rmat

ion

on th

e ac

cide

nt p

roce

sses

, the

re

lativ

e lik

elih

ood

of th

e do

min

ant a

ccid

ent s

eque

nces

, and

th

e as

soci

ated

ope

rato

r act

ions

requ

ired

to p

reve

nt o

r m

itiga

te c

ore

dam

age.

Sim

ilarly

, the

rela

tive

cons

eque

nces

of

var

ious

ope

rato

r err

ors a

nd th

e PS

A-p

redi

cted

cha

nce

of

failu

re c

an b

e us

ed to

sele

ct th

ose

actio

ns th

at w

ould

be

nefit

from

em

phas

ized

trai

ning

. Int

rodu

ctio

n of

SA

MG

s ne

cess

itate

s to

mak

e op

erat

ors u

nder

stan

d se

vere

acc

iden

t sc

enar

ios.

Des

crip

tion

of

dom

inan

t ASs

for

CD

F AV

E and

LE

RF A

VE i

n w

hich

H

Es p

lay

a si

gnifi

cant

role

, ris

k im

porta

nce

mea

sure

s (e.

g. F

-V

and

RA

W) o

f HEs

an

d as

soci

ated

SS

Cs, Δ

CD

F AV

E, Δ

LER

F AV

E

Des

crip

tion

of d

omin

ant a

ccid

ent s

eque

nces

and

ope

rato

r act

ions

to im

plem

ent E

OPs

, rec

over

y ac

tions

, and

SA

MG

s (re

quire

s Lev

el 2

PSA

) with

hig

h ris

k im

porta

nce

are

suff

icie

nt to

enh

ance

tra

inin

g. If

adv

ance

d tra

inin

g is

to b

e ev

alua

ted

as a

cha

nge

to th

e ris

k pr

ofile

, the

HR

A tr

eatm

ent

mus

t be

capa

ble

of m

easu

ring

the

affe

cted

cha

nges

, and

cha

nge

in ri

sk m

etric

s use

to e

valu

ate

the

sign

ifica

nce

and

acce

ptab

ility

of t

he p

ropo

sed

chan

ge.

An

exam

ple

is to

sele

ct d

omin

ant a

ccid

ent s

eque

nces

from

the

PSA

and

incl

ude

som

e of

thes

e in

th

e op

erat

or si

mul

ator

trai

ning

.

3.3.

2 Im

prov

emen

t of m

aint

enan

ce p

erso

nnel

trai

ning

pr

ogra

m

Trai

ning

of m

aint

enan

ce st

aff i

s enh

ance

d ba

sed

on

insi

ghts

and

info

rmat

ion

from

the

PSA

. Foc

us is

on

pote

ntia

l ris

k si

gnifi

cant

impa

cts o

f mai

nten

ance

act

iviti

es,

such

as c

omm

on c

ause

failu

re a

nd m

aint

enan

ce-in

duce

d fa

ilure

of m

ultip

le sy

stem

trai

ns.

This

resu

lts in

an

incr

ease

d fo

cus o

n ris

k-si

gnifi

cant

SSC

s and

on

risk-

sign

ifica

nt fu

nctio

ns a

nd fa

ilure

mod

es th

at m

ust b

e ad

dres

sed

in th

e m

aint

enan

ce p

rogr

am a

s wel

l as

oppo

rtuni

ties t

o op

timiz

e m

aint

enan

ce ta

sks t

hat a

re n

ot

sign

ifica

nt to

risk

man

agem

ent.

Ris

k im

porta

nce

mea

sure

s (e.

g. F

-V,

RA

W) o

f aff

ecte

d SS

Cs,

pre-

acci

dent

H

Es, a

nd b

asic

ev

ents

dea

ling

with

m

aint

enan

ce a

nd

CC

Fs, Δ

CD

F AV

E, Δ

LER

F AV

E

The

risk

impo

rtanc

e m

easu

res a

re u

sed

to ra

nk c

ompo

nent

mai

nten

ance

una

vaila

bilit

ies a

nd p

re-

acci

dent

hum

an e

rror

s and

ass

ocia

ted

com

pone

nt fa

ilure

s tha

t cou

ld b

e in

fluen

ced

by m

aint

enan

ce

prog

ram

cha

nges

; cha

nge

in ri

sk m

etric

s use

d to

eva

luat

e th

e si

gnifi

canc

e an

d ac

cept

abili

ty o

f the

pr

opos

ed c

hang

e to

trai

ning

.

An

exam

ple

is to

trai

n th

e m

aint

enan

ce p

erso

nnel

on

the

SSC

s in

the

scop

e of

the

mai

nten

ance

pr

ogra

m th

at a

re m

ost a

nd le

ast r

isk

sign

ifica

nt a

nd w

hich

SSC

failu

re m

odes

shou

ld g

et th

e m

ost

prio

rity.

Ano

ther

exa

mpl

e is

to u

se P

SA in

sigh

ts to

hel

p m

anag

e th

e m

aint

enan

ce b

ackl

og.

149

Brie

f des

crip

tion

of P

SA a

pplic

atio

n PS

A re

sults

and

m

etric

s for

use

in

deci

sion

mak

ing9

Com

men

ts o

n ho

w P

SA m

odel

s can

be

used

to su

ppor

t app

licat

ion

and

exam

ples

3.3.

3 Im

prov

emen

t of p

lant

man

agem

ent t

rain

ing

prog

ram

The

appl

icat

ion

incl

udes

ade

quat

e co

mm

unic

atio

n of

the

tech

niqu

es, a

pplic

atio

ns a

nd im

plic

atio

ns o

f PSA

to p

lant

m

anag

emen

t to

deve

lop

an in

tegr

al u

nder

stan

ding

in te

rms

of m

anag

emen

t res

pons

ibili

ties.

Furth

erm

ore,

pla

nt

man

agem

ent i

s ulti

mat

ely

resp

onsi

ble

for d

ecis

ions

take

n w

ithin

the

fram

ewor

k of

SA

MG

s. Th

is re

quire

s a g

ood

unde

rsta

ndin

g of

impo

rtant

seve

re a

ccid

ent s

cena

rios,

thei

r fr

eque

ncie

s and

con

sequ

ence

s, as

wel

l as t

he re

latio

nshi

p be

twee

n pl

ant d

esig

n an

d op

erat

iona

l fea

ture

s tha

t im

pact

th

e PS

A re

sults

.

CD

F AV

E, LE

RF A

VE,

QH

Os,

risk

impo

rtanc

e m

easu

res o

f all

mod

elle

d ev

ents

, de

scrip

tion

of

dom

inan

t AS,

risk

in

sigh

ts

The

PSA

resu

lts a

nd a

det

aile

d qu

alita

tive

sum

mar

y of

the

resu

lts a

nd a

ssoc

iate

d ris

k in

sigh

ts a

nd

risk

impo

rtanc

e of

all

mod

elle

d SS

Cs a

nd e

vent

s are

nee

ded

in th

is a

pplic

atio

n to

add

risk

-in

form

ed in

sigh

ts to

the

safe

ty c

ultu

re. I

n ad

ditio

n, th

e pl

ant m

anag

emen

t’s a

ctiv

e pa

rtici

patio

n in

al

l ris

k-in

form

ed a

pplic

atio

n w

ould

bui

ld a

n aw

aren

ess o

f how

to m

anag

e th

e ris

ks. H

ow w

ell t

he

PSA

mod

el re

flect

s the

as-

built

and

as-

oper

ated

pla

nt n

eces

sary

for t

he m

anag

emen

t to

have

co

nfid

ence

in th

e PS

A re

sults

, is o

ne o

f the

mos

t im

porta

nt a

ttrib

utes

for t

his a

pplic

atio

n. T

his

mus

t be

acco

mpa

nied

by

a ba

sic

cour

se in

PSA

con

cept

s and

met

hods

so th

at th

e re

sults

can

be

inte

rpre

ted

prop

erly

.

A st

rikin

g ex

ampl

e is

to im

prov

e th

e sa

fety

cul

ture

and

to e

ngag

e th

e pl

ant m

anag

ers i

n co

nsid

erat

ion

and

man

agin

g th

e ris

k of

acc

iden

ts.

3.4.

Ris

k-ba

sed

conf

igur

atio

n co

ntro

l/ R

isk

Mon

itors

3.4.

1 C

onfig

urat

ion

plan

ning

(e.g

. sup

port

for

plan

t m

aint

enan

ce a

nd te

st a

ctiv

ities

)

Not

e: D

iffer

ent c

ombi

natio

ns o

f equ

ipm

ent c

onfig

urat

ion,

te

sts a

nd m

aint

enan

ce a

ctiv

ities

will

resu

lt in

diff

eren

t le

vels

of r

isk.

The

mai

n be

nefit

of r

isk-

info

rmed

co

nfig

urat

ion

cont

rol i

s the

redu

ctio

n of

risk

pea

ks a

nd th

e co

ntro

l of t

he c

umul

ativ

e, o

r ave

rage

, ris

k. It

hel

ps to

en

sure

that

, as f

ar a

s pos

sibl

e, th

e pl

ant d

oes n

ot e

nter

cr

itica

l, hi

gh-r

isk

situ

atio

ns, t

he p

erio

ds o

f inc

reas

ed ri

sk

are

min

imiz

ed, a

nd th

at o

ther

risk

sign

ifica

nt

conf

igur

atio

ns a

re a

void

ed. T

here

are

two

mai

n ta

sks i

n th

e ris

k-ba

sed

conf

igur

atio

n co

ntro

l: Ta

sk (1

) - ri

sk p

lann

ing

and

Task

(2) -

risk

ass

essm

ent a

nd fo

llow

-up

(see

Item

3.

4.2

on th

e la

tter)

.

This

app

licat

ion

is d

ealin

g w

ith T

ask

(1).

Ris

k pl

anni

ng is

a

forw

ard-

look

ing

appl

icat

ion

of P

SA a

nd it

con

sist

s of

supp

ortin

g th

e pr

epar

atio

n, p

lann

ing

and

sche

dulin

g of

pl

ant a

ctiv

ities

and

com

pone

nt c

onfig

urat

ions

. Thi

s ap

plic

atio

n ca

n be

per

form

ed w

ith a

n on

-line

or o

ff-li

ne

PSA

mod

el.

CD

F{t}

, LER

F{t}

, IC

CD

P, IC

LER

P,

risk

impo

rtanc

e m

easu

res o

f SSC

s as

a fu

nctio

n of

tim

e

The

PSA

mod

el u

sed

for t

he ‘b

ase

case

PSA

’ mus

t be

mod

ified

to h

ave

the

poss

ibili

ty to

el

imin

ate

time

aver

aged

mai

nten

ance

una

vaila

bilit

ies a

nd a

vera

ge m

odel

s for

alte

rnat

ive

conf

igur

atio

ns, a

nd re

plac

e th

em w

ith b

inar

y (‘

On’

, ‘O

ff’)

type

mod

els t

hat m

ake

the

CD

F an

d LE

RF

resu

lts d

epen

dent

on

spec

ific

plan

t con

figur

atio

ns a

nd e

quip

men

t out

-of-

serv

ice

cond

ition

s to

be

eval

uate

d. M

odel

ling

sim

plifi

catio

n to

trea

t sym

met

ric tr

ains

with

shar

ed b

asic

eve

nts m

ust

be e

xpan

ded

to m

odel

eac

h tra

in e

xplic

itly.

Tru

ncat

ion

issu

es m

ust b

e re

solv

ed so

the

PSA

resu

lts

are

valu

ed w

ith a

ny c

ombi

natio

n of

equ

ipm

ent t

o be

eva

luat

ed a

ssum

ed to

be

out o

f ser

vice

. A

m

eans

of s

peci

fyin

g pl

ant s

tate

cha

nges

, as a

func

tion

of ti

me

into

the

Ris

k M

onito

r mus

t be

prov

ided

. The

risk

impo

rtanc

e m

easu

res i

n th

is a

pplic

atio

n ar

e us

ed to

dev

elop

‘ret

urn

to se

rvic

e’

prio

ritie

s and

‘rem

ain

in se

rvic

e’ p

riorit

ies f

or e

ach

SSC

. Int

erfa

ce w

ith sc

hedu

ling

softw

are

is

help

ful.

The

PSA

mod

el sh

ould

be

capa

ble

of p

redi

ctin

g th

e te

mpo

ral c

hang

es in

initi

atin

g ev

ent

freq

uenc

ies,

com

pone

nt u

nava

ilabi

litie

s due

mai

nten

ance

and

oth

er c

onfig

urat

ion

chan

ges o

n C

DF

and

LER

F.

In c

ase

the

appl

icat

ion

is a

imed

to a

sses

s the

risk

ass

ocia

ted

with

tran

sitio

ns b

etw

een

diff

eren

t po

wer

mod

es, t

he P

SA m

odel

shou

ld in

corp

orat

e th

e IE

s and

syst

em c

onfig

urat

ions

for d

iffer

ent

plan

t ope

ratin

g st

ates

with

in th

e sc

ope

of p

ower

PSA

(e.g

. ope

ratio

n w

ith o

ne tu

rbin

e, o

pera

tion

with

no

turb

ine

abov

e 2%

of n

omin

al p

ower

, con

nect

ion

to re

serv

e ex

tern

al g

rid).

An

exam

ple

is th

e us

e of

a R

isk

Mon

itor t

ool t

o ev

alua

te th

e tim

e-de

pend

ent r

isk

prof

ile fo

r a

futu

re ti

me

perio

d, in

whi

ch a

serie

s of p

lant

con

figur

atio

n ch

ange

s is b

eing

pla

nned

.

150

Brie

f des

crip

tion

of P

SA a

pplic

atio

n PS

A re

sults

and

m

etric

s for

use

in

deci

sion

mak

ing9

Com

men

ts o

n ho

w P

SA m

odel

s can

be

used

to su

ppor

t app

licat

ion

and

exam

ples

3.4.

2 R

eal t

ime

conf

igur

atio

n as

sess

men

t and

con

trol

(r

espo

nse

to e

mer

ging

con

ditio

ns)

This

app

licat

ion

is d

ealin

g w

ith T

ask

(2) d

iscu

ssed

abo

ve

in It

em 3

.4.1

.

Ris

k as

sess

men

t and

follo

w-u

p in

volv

es th

e on

line

use

of

the

PSA

by

plan

t per

sonn

el in

ord

er to

kee

p th

e ris

k du

e to

ac

tual

con

figur

atio

ns, p

lant

act

iviti

es a

nd u

nant

icip

ated

ev

ents

at a

n ac

cept

able

leve

l.

CD

F{t}

, LER

F{t}

, IC

CD

P, IC

LER

P,

risk

impo

rtanc

e m

easu

res o

f SSC

s as

a fu

nctio

n of

tim

e

Use

of P

SA m

odel

is e

ssen

tially

the

sam

e as

des

crib

ed in

Item

3.4

.1 e

xcep

t for

the

time

fram

e ov

er w

hich

the

risk

to b

e ev

alua

ted

is d

iffer

ent (

i.e. c

ondi

tione

d by

the

dura

tion

of e

mer

ging

co

nditi

ons)

.

An

exam

ple

is th

e us

e of

a R

isk

Mon

itor t

ool t

o pe

rfor

m p

ost m

orte

m e

valu

atio

n of

the

time

depe

nden

t ris

k pr

ofile

for a

pre

viou

s tim

e pe

riod

in w

hich

a se

ries o

f pla

nt c

onfig

urat

ion

chan

ges

has o

ccur

red.

This

app

licat

ion

may

als

o re

quire

revi

ewin

g th

e pe

rfor

man

ce o

f the

mon

itorin

g to

ols t

o en

sure

th

ey a

re a

ble

to m

easu

re th

e ris

k im

pact

s of a

ll ac

tiviti

es th

at w

ere

expe

rienc

ed.

3.4.

3 E

xem

ptio

ns to

TS

and

just

ifica

tion

for

cont

inue

d op

erat

ion

As a

com

preh

ensi

ve to

ol d

escr

ibin

g th

e ris

k as

soci

ated

w

ith a

par

ticul

ar p

lant

con

figur

atio

n, th

e PS

A c

an p

rovi

de

usef

ul su

ppor

t to

TS e

xem

ptio

n ju

stifi

catio

ns a

nd/o

r to

prop

osal

s for

miti

gatio

n or

com

pens

ator

y m

easu

res,

or to

ju

stify

the

rele

vanc

e of

thes

e m

easu

res.

ΔC

DF A

VE,

ΔLE

RF A

VE,

ICC

DP,

IC

LER

P

The

scop

e of

this

app

licat

ion

is n

orm

ally

con

fined

to a

subs

et o

f SSC

s tha

t are

cur

rent

ly in

clud

ed

in th

e ba

se c

ase

PSA

mod

el; o

ther

wis

e th

e PS

A is

upd

ated

to in

corp

orat

e al

l affe

cted

SSC

s ex

plic

itly.

The

PSA

mod

el m

ust e

xplic

itly

mod

el th

e ar

eas a

ffect

ed b

y th

e TS

cha

nge.

The

ch

ange

in ri

sk m

etric

s is u

sed

to e

valu

ate

the

risk

sign

ifica

nce

and

acce

ptab

ility

of t

he p

ropo

sed

chan

ge.

This

app

licat

ion

is d

ealin

g w

ith te

mpo

rary

cha

nges

and

hen

ce th

e de

cisi

on c

riter

ia m

ay b

e le

ss

rest

rictiv

e th

an fo

r the

cas

e of

per

man

ent c

hang

es b

ecau

se o

f one

-tim

e na

ture

of t

he e

xem

ptio

n.

An

exam

ple

is th

e fa

ilure

of a

n em

erge

ncy

feed

wat

er p

ump

shaf

t tha

t req

uire

s a w

eek

to re

pair

and

a ju

stifi

catio

n to

rela

x th

e 72

hou

r AO

T on

an

one-

time

basi

s whi

le ta

king

com

pens

atin

g m

easu

res t

o m

inim

ize

the

risk

impa

cts.

3.4.

4 D

ynam

ic r

isk-

info

rmed

TS

Typi

cally

, cla

ssic

al T

Ss c

onsi

st o

f a ri

gid

fram

ewor

k of

pr

escr

iptio

ns fo

r ind

ivid

ual e

quip

men

t and

syst

ems

invo

lvin

g fix

ed g

race

tim

es, f

or e

xam

ple.

The

pur

pose

of

this

rigi

d fr

amew

ork

is to

kee

p pl

ant f

eatu

res w

ithin

the

licen

sing

bas

is fo

r a re

ason

ably

larg

e fr

actio

n of

tim

e.

Dyn

amic

risk

-info

rmed

TS

rela

xes s

ome

of th

is ri

gidi

ty

base

d on

the

inte

gral

vie

w o

f a P

SA. A

OTs

are

cal

cula

ted

for e

ach

plan

t sta

te ta

king

into

acc

ount

the

com

plet

e pi

ctur

e of

pla

nt c

onfig

urat

ion

and

equi

pmen

t out

-of-

serv

ice

com

bina

tions

. In

som

e ca

ses t

he c

alcu

late

d A

OT

may

be

shor

ter t

han

the

stan

dard

fixe

d te

chni

cal s

peci

ficat

ion

vers

ion

and

in so

me

case

s a lo

nger

AO

T is

just

ified

, but

in

all c

ases

a c

aref

ul e

valu

atio

n is

mad

e at

all

times

of t

he

com

plet

e pi

ctur

e of

pla

nt c

onfig

urat

ions

incl

udin

g sa

fety

an

d no

n-sa

fety

rela

ted

SSC

s.

CD

F{t}

, LER

F{t}

, IC

CD

P, IC

LER

P Th

is a

pplic

atio

n is

sim

ilar t

o Ite

m 3

.4.3

exc

ept t

hat n

ew A

OTs

are

not

fixe

d in

the

tech

nica

l sp

ecifi

catio

n do

cum

ent b

ut d

ynam

ical

ly c

alcu

late

d ba

sed

on th

e st

atus

of a

ll SS

Cs i

n th

e pl

ant a

nd

a tim

e de

pend

ent R

isk

Mon

itor.

An

exam

ple

of th

is is

a R

isk

Mon

itor t

hat c

alcu

late

s a n

ew A

OT

for e

ach

adve

rse

conf

igur

atio

n ch

ange

mad

e by

the

plan

t.

151

Brie

f des

crip

tion

of P

SA a

pplic

atio

n PS

A re

sults

and

m

etric

s for

use

in

deci

sion

mak

ing9

Com

men

ts o

n ho

w P

SA m

odel

s can

be

used

to su

ppor

t app

licat

ion

and

exam

ples

4. P

ER

MA

NE

NT

CH

AN

GE

S T

O T

HE

OPE

RA

TIN

G P

LA

NT

4.1

Pla

nt c

hang

es

4.1.

1 N

PP u

pgra

des,

back

-fitt

ing

activ

ities

and

pla

nt

mod

ifica

tions

Iden

tific

atio

n of

wea

knes

ses a

nd e

ffec

tive

area

s for

im

prov

emen

t in

plan

t des

ign

and

oper

atio

nal f

eatu

res i

n vi

ew o

f pla

nt ri

sk. A

sses

smen

t may

incl

ude

inve

stig

atio

n of

var

iant

s and

of e

xplo

rato

ry o

ptio

ns. P

SA a

rgum

ents

are

us

ed to

supp

ort t

he se

lect

ion,

des

ign,

impl

emen

tatio

n,

just

ifica

tion,

and

lice

nsin

g of

pla

nt u

pgra

des.

Ris

k im

porta

nce

mea

sure

s (e.

g. F

-V,

RA

W) o

f aff

ecte

d SS

Cs a

nd h

uman

ac

tions

, ΔC

DF A

VE,

ΔLE

RF A

VE

Impo

rtanc

e m

easu

res f

rom

‘bas

e ca

se P

SA’ a

re u

sed

to h

elp

prio

ritiz

e ca

ndid

ate

desi

gn c

hang

es;

chan

ge in

CD

F an

d LE

RF

is u

sed

to ju

stify

acc

epta

ble

risk

impa

cts a

nd to

det

erm

ine

risk

sign

ifica

nce.

The

leve

l of d

etai

l of t

he P

SA m

odel

in th

e ar

eas a

ffec

ted

by th

e de

sign

cha

nges

may

be

gre

ater

than

that

for t

he re

st o

f the

pla

nt; a

mor

e si

mpl

ified

con

serv

ativ

e tre

atm

ent o

f oth

er

parts

of t

he m

odel

acc

epta

ble.

Dat

a fo

r new

add

ition

al e

quip

men

t may

not

be

avai

labl

e; th

eref

ore,

tre

atm

ent o

f suc

h eq

uipm

ent i

n PS

A m

odel

shou

ld b

e ju

stifi

ed.

A ty

pica

l exa

mpl

e fo

r suc

h ap

plic

atio

n is

the

intro

duct

ion

of th

e pr

imar

y fe

ed a

nd b

leed

feat

ure

in

a PW

R N

PP, w

hich

wou

ld in

clud

e ha

rdw

are

upgr

ades

such

as i

nsta

llatio

n of

relie

f val

ves w

hich

ar

e qu

alifi

ed fo

r fee

d an

d bl

eed

and

the

elab

orat

ion

and

impl

emen

tatio

n of

ass

ocia

ted

proc

edur

es.

4.1.

2 L

ifetim

e ex

tens

ion

The

appl

icat

ion

is o

ften

refe

rred

as a

sub-

case

of t

he

perio

dic

safe

ty re

view

with

the

cons

ider

atio

n of

agi

ng

effe

cts b

eyon

d th

e de

sign

life

time.

Invo

lves

mod

ellin

g of

ag

ing

effe

cts i

n PS

A. (

See

also

Item

1.2

.)

CD

F AV

E, LE

RF A

VE,

CD

P, L

ERP,

QH

Os,

risk

impo

rtanc

e m

easu

res o

f all

SSC

s and

IEs,

prim

ary

cont

ribut

ors

to ri

sk

A fu

ll sc

ope

PSA

is n

eede

d to

add

ress

the

appl

icat

ion

in c

ompl

ete

and

cons

iste

nt m

anne

r. Th

e ke

y is

sue

in th

is a

pplic

atio

n is

the

poss

ible

impa

ct o

f agi

ng p

heno

men

a an

d co

mpo

nent

life

time

cons

ider

atio

ns b

eyon

d th

e de

sign

life

time

on th

e ov

eral

l ris

k m

etric

s. M

odel

ling

the

agin

g ph

enom

ena

is in

the

expl

orat

ory

stag

e; in

prin

cipl

e th

e PS

A sh

ould

be

capa

ble

of e

stim

atin

g or

bo

undi

ng th

e po

ssib

le e

ffec

ts o

f agi

ng o

n pa

ssiv

e co

mpo

nent

s tha

t are

not

nor

mal

ly m

aint

aine

d or

re

plac

ed. T

ime

trend

ana

lyse

s in

rela

tion

to IE

freq

uenc

ies,

equi

pmen

t fai

lure

rate

s, ca

ble

mat

eria

l pr

oper

ties,

etc.

supp

ort t

he a

pplic

atio

n. T

he a

naly

sis r

esul

ts p

rovi

de a

dditi

onal

info

rmat

ion

for

regu

lato

rs w

hile

lice

nsin

g th

e lif

etim

e ex

tens

ion.

An

exam

ple

is th

e ev

alua

tion

of th

e in

crea

se in

risk

due

to a

ging

of p

lant

equ

ipm

ent p

ast t

he

desi

gn li

fetim

e. T

ypic

al e

quip

men

t of m

ajor

inte

rest

are

: re

acto

r ves

sel,

stea

m g

ener

ator

, pip

ing,

et

c.

152

Brie

f des

crip

tion

of P

SA a

pplic

atio

n PS

A re

sults

and

m

etric

s for

use

in

deci

sion

mak

ing9

Com

men

ts o

n ho

w P

SA m

odel

s can

be

used

to su

ppor

t app

licat

ion

and

exam

ples

4.2

Tech

nica

l spe

cific

atio

n ch

ange

s

4.2.

1 D

eter

min

atio

n an

d ev

alua

tion

of c

hang

es to

al

low

ed o

utag

e tim

e an

d ch

ange

s to

requ

ired

TS

actio

ns

Req

uire

d ac

tions

in T

S ha

ve b

een

typi

cally

der

ived

on

a cl

assi

cal e

ngin

eerin

g ba

sis.

This

invo

lves

, for

exa

mpl

e,

mod

ified

surv

eilla

nce

activ

ities

to c

ompe

nsat

e fo

r red

uced

av

aila

bilit

y of

equ

ipm

ent.

Such

item

s can

be

re-e

valu

ated

ba

sed

on th

e PS

A a

nd c

hang

ed e

vent

ually

acc

ordi

ng to

ris

k-si

gnifi

canc

e. F

ocus

is o

n th

e ris

k im

pact

due

to th

e A

OT

perio

d. T

his r

equi

res a

sses

smen

t of t

hree

type

s of

risks

: (a)

inst

anta

neou

s (co

nditi

onal

) ris

k w

hile

the

com

pone

nt is

in m

aint

enan

ce; (

b) c

umul

ativ

e (in

tegr

ated

) ris

k ov

er th

e A

OT

perio

d; (c

) ave

rage

risk

ove

r a lo

ng

perio

d (e

.g. y

early

), ta

king

into

acc

ount

the

freq

uenc

y of

m

aint

enan

ce p

erfo

rmed

on

a co

mpo

nent

. Opt

imum

AO

T m

ay in

volv

e tra

de-o

ffs b

etw

een

exte

nded

equ

ipm

ent

unav

aila

bilit

y du

ring

pow

er o

pera

tion

and

unav

aila

bilit

y of

th

e sa

me

equi

pmen

t dur

ing

shut

dow

n co

nditi

ons.

The

ultim

ate

goal

is to

find

the

optim

um A

OT

for e

ach

SSC

co

vere

d in

the

tech

nica

l spe

cific

atio

n w

ith re

spec

t to

how

it

cons

train

s pla

nt o

pera

tion

stat

es a

nd h

ow it

is u

sed

to

man

age

risks

of e

quip

men

t bei

ng o

ut o

f ser

vice

.

In a

dditi

on, t

he P

SA m

ay b

e us

ed to

supp

ort t

he

optim

izat

ion

of m

aint

enan

ce ta

sks w

ith re

spec

t to

whe

ther

th

ey m

ust b

e do

ne d

urin

g ou

tage

s or w

heth

er o

n-lin

e m

aint

enan

ce is

app

ropr

iate

.

Ris

k im

porta

nce

mea

sure

s of a

ffec

ted

SSC

s, Δ

CD

F AV

E, Δ

LER

F AV

E, IC

CD

P,

ICLE

RP

The

scop

e of

this

app

licat

ion

is n

orm

ally

con

fined

to a

subs

et o

f SSC

s tha

t are

cur

rent

ly in

clud

ed

in th

e ‘b

ase

case

PSA

’ mod

el; o

ther

wis

e th

e PS

A is

upd

ated

to in

corp

orat

e al

l affe

cted

SSC

s ex

plic

itly.

The

leve

l of d

etai

l of t

he P

SA m

odel

in th

e ar

eas a

ffect

ed b

y th

e A

OT

chan

ges m

ay b

e gr

eate

r tha

n th

at fo

r the

rest

of t

he p

lant

. The

PSA

mod

el m

ust e

xplic

itly

mod

el th

e m

aint

enan

ce

unav

aila

bilit

y fo

r all

SSC

s who

se A

OTs

hav

e be

en c

hang

ed.

The

mod

el sh

all b

e ca

pabl

e to

pro

perly

refle

ct a

ll ef

fect

s of s

yste

m/c

ompo

nent

una

vaila

bilit

y:

- se

tting

com

pone

nts/

syst

em to

una

vaila

ble

stat

e;

- ba

lanc

ed e

ffect

of u

nava

ilabi

lity

of p

artic

ular

redu

ndan

t tra

in/c

ompo

nent

(sym

met

ric

mod

el, e

.g. s

team

line

rupt

ure

is re

pres

ente

d by

rupt

ure

on S

G1,

so th

e ef

fect

of

the

unav

aila

bilit

y of

SG

1 is

olat

ion

valv

e is

ove

rest

imat

ed a

nd th

e ef

fect

of u

nava

ilabi

litie

s of

othe

r equ

ival

ent i

sola

tion

valv

es is

und

eres

timat

ed).

The

chan

ges i

n ris

k m

etric

s are

use

d to

eva

luat

e th

e ris

k si

gnifi

canc

e an

d ac

cept

abili

ty o

f the

pr

opos

ed c

hang

e an

d th

e in

crem

enta

l ris

k m

etric

s are

use

d to

eva

luat

e th

e ac

cept

abili

ty o

f the

new

pr

opos

ed A

OT.

Exam

ple:

One

of t

he m

ost c

omm

on e

xam

ples

is a

n ex

tens

ion

of th

e ED

G A

OTs

to p

erm

it on

-line

ove

rhau

ls

of th

e co

mpo

nent

s dur

ing

pow

er o

pera

tion.

153

Brie

f des

crip

tion

of P

SA a

pplic

atio

n PS

A re

sults

and

m

etric

s for

use

in

deci

sion

mak

ing9

Com

men

ts o

n ho

w P

SA m

odel

s can

be

used

to su

ppor

t app

licat

ion

and

exam

ples

4.2.

2 R

isk-

info

rmed

opt

imis

atio

n of

TS

The

tech

nica

l spe

cific

atio

ns d

efin

e lim

its a

nd c

ondi

tions

fo

r ope

ratio

n, te

stin

g, a

nd m

aint

enan

ce a

ctiv

ities

as a

way

to

ass

ure

that

the

plan

t is o

pera

ted

safe

ly. F

rom

tim

e to

tim

e, th

e pl

ant o

pera

tor m

ay n

eed

a TS

exe

mpt

ion

due

to

oper

atio

nal b

urde

ns a

nd c

onst

rain

ts. T

his a

pplic

atio

n is

us

ed to

opt

imis

e TS

pro

visi

ons.

Bes

ide

risk

insi

ghts

, oth

er n

ot ri

sk-b

ased

par

amet

ers m

ay

have

to b

e us

ed a

s con

stra

ints

in th

e op

timis

atio

n pr

oces

s.

Ris

k im

porta

nce

mea

sure

s of a

ffec

ted

SSC

s, Δ

CD

F AV

E, Δ

LER

F AV

E, IC

CD

P,

ICLE

RP

This

app

licat

ion

is si

mila

r to

Item

4.2

.1 d

epen

ding

on

the

natu

re o

f the

TS

actio

n to

be

chan

ged.

This

app

licat

ion

invo

lves

the

defin

ition

of c

onst

rain

ing

para

met

ers b

esid

e ris

k m

etric

s to

assu

re

effic

ient

resu

lts (e

.g. c

ost-b

enef

it co

rrel

atio

ns).

An

exam

ple

is a

pro

posa

l to

rem

ove

as te

chni

cal s

peci

ficat

ion

requ

irem

ent t

o te

st a

n op

erab

le

EDG

eve

ry h

our w

hile

a re

dund

ant E

DG

trai

n is

out

of s

ervi

ce.

4.2.

3 D

eter

min

atio

n an

d ev

alua

tion

of c

hang

es to

su

rvei

llanc

e te

st in

terv

als

PSA

bas

ed e

valu

atio

n of

surv

eilla

nce

test

inte

rval

s (ST

Is)

cons

ider

s the

risk

from

una

vaila

bilit

y du

e to

und

etec

ted

failu

res,

and

the

risk

from

una

vaila

bilit

y du

e to

test

s and

te

st in

duce

d fa

ilure

s. Th

e go

al is

to o

ptim

ize

the

STIs

with

re

spec

t to

thei

r im

pact

on

equi

pmen

t rel

iabi

lity

and

how

th

ese

test

s im

pact

the

cost

of o

pera

tions

. H

uman

err

ors

durin

g ST

Is th

at m

ay h

ave

an a

dver

se im

pact

on

safe

ty, f

or

exam

ple

by le

adin

g to

pla

nt tr

ips a

nd in

itiat

ing

even

ts

norm

ally

is c

onsi

dere

d in

dec

idin

g th

is o

ptim

izat

ion.

Ris

k im

porta

nce

mea

sure

s (e.

g. F

-V,

RA

W) o

f aff

ecte

d SS

Cs, Δ

CD

F AV

E, Δ

LER

F AV

E

The

scop

e of

this

app

licat

ion

is n

orm

ally

con

fined

to a

subs

et o

f SSC

s tha

t are

cur

rent

ly in

clud

ed

in th

e ‘b

ase

case

PSA

’ mod

el; o

ther

wis

e th

e PS

A is

upd

ated

to in

corp

orat

e th

e af

fect

ed S

SCs

expl

icitl

y if

they

can

be

used

in a

ccid

ent m

itiga

tion.

The

leve

l of d

etai

l of t

he P

SA m

odel

in th

e ar

eas a

ffect

ed b

y th

e ST

I cha

nges

may

be

grea

ter t

han

that

for t

he re

st o

f the

pla

nt.

The

PSA

m

odel

mus

t exp

licitl

y m

odel

test

una

vaila

bilit

y of

the

SSC

and

pro

vide

a c

apab

ility

to p

redi

ct th

e im

pact

of c

hang

es to

the

STI o

n ea

ch a

ffect

ed c

ompo

nent

una

vaila

bilit

y du

e to

a ra

ndom

failu

re.

R

isk

impo

rtanc

e m

easu

res c

an b

e us

ed to

prio

ritiz

e an

d ra

nk th

e ca

ndid

ates

for S

TI c

hang

e. T

he

chan

ge in

risk

met

rics i

s use

d to

eva

luat

e th

e ris

k si

gnifi

canc

e an

d ac

cept

abili

ty o

f the

pro

pose

d ch

ange

and

the

incr

emen

tal r

isk

met

rics a

re u

sed

to e

valu

ate

the

acce

ptab

ility

of t

he n

ew p

ropo

sed

STI.

An

unde

rsta

ndin

g of

how

hum

an e

rror

s dur

ing

test

ing

cont

ribut

e to

initi

atin

g ev

ent

freq

uenc

ies a

nd c

ompo

nent

failu

res i

s nee

ded

to b

alan

ce th

e po

sitiv

e an

d ne

gativ

e as

pect

s of

surv

eilla

nce

test

ing.

Una

vaila

bilit

y of

equ

ipm

ent d

ue to

hum

an e

rror

s to

prop

erly

rest

ore

norm

al

alig

nmen

ts a

fter t

estin

g is

to b

e ta

ken

into

acc

ount

. If i

t is k

now

n th

at a

test

may

lead

to a

hig

her

prob

abili

ty o

f an

initi

atin

g ev

ent (

initi

atin

g ev

ent f

requ

ency

is re

late

d to

test

freq

uenc

y) th

en th

is

rela

tions

hip

mus

t be

take

n in

to a

ccou

nt if

the

test

freq

uenc

y is

cha

nged

.

A ty

pica

l exa

mpl

e is

the

test

of s

afet

y in

ject

ion

train

s by

runn

ing

one

train

via

a fu

ll ca

paci

ty

bypa

ss li

ne b

ack

to th

e su

ppre

ssio

n po

ol (B

WR

) or b

ack

to th

e bo

rate

d w

ater

stor

age

tank

(PW

R).

Typi

cally

, the

se s

yste

ms a

re a

utom

atic

ally

reco

nfig

ured

from

the

test

con

figur

atio

n an

d st

arte

d w

hen

a re

al d

eman

d ha

ppen

s. Th

us, w

hen

incr

easi

ng te

st fr

eque

ncy,

ther

e is

a tr

ade-

off b

etw

een

redu

ced

com

pone

nt u

nava

ilabi

lity,

incr

ease

d un

avai

labi

lity

of e

quip

men

t dur

ing

the

test

due

to

real

ignm

ent o

f val

ves,

actu

atio

n of

con

trol c

ircui

ts, o

peni

ng o

f AC

or D

C c

ircui

t bre

aker

s, et

c.

154

Brie

f des

crip

tion

of P

SA a

pplic

atio

n PS

A re

sults

and

m

etric

s for

use

in

deci

sion

mak

ing9

Com

men

ts o

n ho

w P

SA m

odel

s can

be

used

to su

ppor

t app

licat

ion

and

exam

ples

4.2.

4 R

isk-

info

rmed

in-s

ervi

ce te

stin

g

Use

of P

SA to

supp

ort t

he IS

T pr

ogra

mm

e, ta

king

into

ac

coun

t the

rela

tive

risk

sign

ifica

nce

of th

e co

mpo

nent

s to

be te

sted

, is i

n th

e fo

cus o

f the

app

licat

ion.

Thi

s app

licat

ion

is n

orm

ally

lim

ited

to p

umps

and

val

ves i

n sa

fety

rela

ted

syst

ems,

but i

n pr

inci

ple

can

be a

pplie

d to

any

com

pone

nt

in th

e IS

T pr

ogra

m. T

he re

lativ

e ris

k si

gnifi

canc

e is

as

sess

ed u

sing

a b

lend

of p

roba

bilis

tic a

nd d

eter

min

istic

m

etho

ds b

efor

e an

y te

st in

terv

al is

cha

nged

and

the

aggr

egat

e im

pact

of t

he c

hang

es is

eva

luat

ed.

This

resu

lts

in re

laxa

tion

of te

stin

g re

quire

men

ts fo

r low

risk

si

gnifi

cant

SSC

s and

SSC

func

tions

and

incr

ease

d re

quire

men

ts fo

r hig

her r

isk

sign

ifica

nt S

SCs y

ield

ing

an

optim

ized

test

ing

prog

ram

.

Ris

k im

porta

nce

mea

sure

s (e.

g. F

-V,

RA

W) o

f aff

ecte

d SS

Cs, Δ

CD

F AV

E, Δ

LER

F AV

E

Ris

k im

porta

nce

mea

sure

s fro

m th

e ba

se P

SA u

sed

to h

elp

prio

ritiz

e ca

ndid

ate

chan

ges t

o IS

T pr

ogra

m fo

r sel

ecte

d pu

mps

and

val

ves,

chan

ge in

CD

F an

d LE

RF

used

to ju

stify

acc

epta

ble

risk

impa

cts a

nd to

det

erm

ine

risk

sign

ifica

nce.

Exp

licit

mod

el o

f tes

t and

mai

nten

ance

un

avai

labi

litie

s and

cap

abili

ty to

pre

dict

or b

ound

the

impa

ct o

f pro

gram

cha

nges

on

com

pone

nt

unav

aila

bilit

y du

e to

a ra

ndom

failu

re a

nd te

st a

nd m

aint

enan

ce u

nava

ilabi

litie

s nee

ded

to su

ppor

t th

is a

pplic

atio

n.

The

PSA

can

be

used

to a

naly

se h

ow a

cha

nge

in te

st in

terv

als a

ffec

ts th

e pl

ant r

isk

taki

ng in

to

acco

unt n

egat

ive

effe

cts s

uch

as p

oten

tially

incr

ease

d fr

eque

ncy

of p

lant

tran

sien

ts, e

.g. t

estin

g st

rate

gy fo

r pre

ssur

izer

relie

f val

ves,

MSI

Vs,

etc.

4.2.

5 R

isk-

info

rmed

in-s

ervi

ce in

spec

tions

The

risk-

info

rmed

in-s

ervi

ce in

spec

tion

(RI-

ISI)

m

etho

dolo

gy c

onsi

sts o

f ran

king

the

elem

ents

for

insp

ectio

n, su

ch a

s wel

ds in

pip

ing

syst

ems,

acco

rdin

g to

th

eir r

isk

sign

ifica

nce

and

deve

lopi

ng th

e in

spec

tion

stra

tegy

(fre

quen

cy, m

etho

d, sa

mpl

e si

ze, e

tc.)

com

men

sura

te w

ith th

eir r

isk

sign

ifica

nce.

It p

rovi

des a

fr

amew

ork

for e

ffec

tive

allo

catio

n of

insp

ectio

n re

sour

ces

and

help

s to

focu

s the

insp

ectio

n ac

tiviti

es w

here

they

are

m

ost n

eede

d. In

add

ition

, an

unde

rsta

ndin

g of

the

mos

t lik

ely

degr

adat

ion

mec

hani

sms i

s dev

elop

ed, w

hich

is u

sed

to fo

cus r

equi

red

insp

ectio

ns to

use

the

mos

t app

ropr

iate

in

spec

tion

met

hods

for t

he a

ntic

ipat

ed d

amag

e m

echa

nism

s.

Com

pone

nt fa

ilure

ra

tes f

or d

iffer

ent

insp

ectio

n st

rate

gies

(e

.g. o

btai

ned

usin

g PF

M o

r Mar

kov

mod

el),

CC

DP,

C

LER

P, Δ

CD

F AV

E, Δ

LER

F AV

E

This

app

licat

ion

is n

orm

ally

lim

ited

to p

ipin

g sy

stem

insp

ectio

ns, b

ut in

prin

cipl

e ca

n be

app

lied

to a

ny p

assi

ve c

ompo

nent

cov

ered

in th

e In

-Ser

vice

Insp

ectio

n (I

SI) p

rogr

am. T

hese

pas

sive

co

mpo

nent

s are

nor

mal

ly n

ot e

xplic

itly

mod

elle

d in

a ‘b

ase

case

PSA

’. H

ence

, spe

cial

ana

lyse

s m

ust b

e pe

rfor

med

to e

stim

ate

com

pone

nt (e

.g. w

eld)

leve

l fai

lure

rate

s as a

func

tion

of le

vel a

nd

type

of i

nspe

ctio

n, a

nd c

onse

quen

ces o

f pip

e fa

ilure

in te

rms o

f the

CC

DP

and

CLE

RP

due

to th

e lo

ss o

f fun

ctio

n an

d se

cond

ary

flood

ing

and

othe

r con

sequ

ence

s of s

yste

m p

ipe

brea

ks.

Thes

e sp

ecia

l ana

lyse

s are

use

d to

dev

elop

risk

impo

rtanc

e m

easu

res o

r alte

rnat

ive

risk

rank

ing

mat

rices

w

hich

are

use

d to

hel

p pr

iorit

ize

cand

idat

e ch

ange

s to

ISI p

rogr

am fo

r sel

ecte

d w

eld

loca

tions

, ch

ange

in C

DF

and

LER

F us

ed to

just

ify a

ccep

tabl

e ris

k im

pact

s and

to d

eter

min

e ris

k si

gnifi

canc

e. T

he b

ase

PSA

mod

el m

ust b

e ca

pabl

e of

supp

ortin

g es

timat

es o

f the

CC

DP

and

CLE

RP

of a

ny a

ssum

ed fa

ilure

mod

e w

ithin

the

scop

e of

the

pipi

ng sy

stem

s sel

ecte

d fo

r the

RI-

ISI p

rogr

am.

To d

ate,

mos

t exa

mpl

es in

volv

e in

-ser

vice

insp

ectio

ns o

f wel

ds in

NPP

pip

ing

syst

ems i

n w

hich

ca

se th

e nu

mbe

r, fr

eque

ncy,

and

met

hod

of n

on-d

estru

ctiv

e ex

amin

atio

n (N

DE)

are

var

ied

to

impr

ove

the

allo

catio

n of

insp

ectio

n re

sour

ces t

o th

e m

ost r

isk

sign

ifica

nt p

ipe

elem

ents

and

to

man

age

the

risk

of in

spec

tions

with

resp

ect t

o pi

pe ru

ptur

es. F

utur

e ex

ampl

es c

over

the

brea

dth

of

SSC

s cur

rent

ly c

over

ed in

ISI p

rogr

ams.

155

Brie

f des

crip

tion

of P

SA a

pplic

atio

n PS

A re

sults

and

m

etric

s for

use

in

deci

sion

mak

ing9

Com

men

ts o

n ho

w P

SA m

odel

s can

be

used

to su

ppor

t app

licat

ion

and

exam

ples

4.3

Est

ablis

hmen

t of g

rade

d Q

A p

rogr

am fo

r SSC

4.3.

1 E

quip

men

t ris

k si

gnifi

canc

e ev

alua

tion

PSA

pro

vide

s the

nec

essa

ry in

sigh

ts th

at a

re u

sed

to

dete

rmin

e th

e re

lativ

e sa

fety

sign

ifica

nce

of p

lant

eq

uipm

ent.

Thes

e pr

obab

ilist

ic in

sigh

ts a

re fo

r exa

mpl

e ut

ilize

d to

hel

p id

entif

y lo

w/h

igh

safe

ty si

gnifi

cant

SSC

s th

at a

re c

andi

date

s for

redu

ctio

ns/im

prov

emen

ts in

QA

tre

atm

ent.

Ris

k im

porta

nce

mea

sure

s of a

ffec

ted

SSC

s (e.

g. F

-V,

RA

W)

Ris

k im

porta

nce

mea

sure

s are

use

d to

cla

ssify

the

risk

and

safe

ty si

gnifi

canc

e of

SSC

s. T

his

info

rmat

ion

is c

onsi

dere

d in

con

nect

ion

with

the

curr

ent s

afet

y cl

assi

ficat

ion

of S

SCs a

nd

asso

ciat

ed Q

A a

nd sp

ecia

l tre

atm

ent r

equi

rem

ents

; thi

s is u

sed

to id

entif

y an

d ra

nk c

andi

date

SS

Cs f

or p

ropo

sed

chan

ge in

QA

and

spec

ial t

reat

men

t req

uire

men

ts.

An

exam

ple

is p

laci

ng a

ll pl

ant S

SCs i

nto

diffe

rent

cat

egor

ies b

ased

on

safe

ty re

late

d cl

assi

ficat

ion

and

risk

impo

rtanc

e us

ing

F-V

and

RA

W ty

pe o

f mea

sure

s.

4.3.

2 E

valu

atio

n of

ris

k im

pact

of c

hang

es to

QA

re

quir

emen

ts

Cha

nges

in Q

A tr

eatm

ent o

f SSC

s are

inve

stig

ated

or

expl

ored

with

in th

e fr

amew

ork

of P

SA.

Ris

k im

porta

nce

mea

sure

s (e.

g. F

-V,

RA

W) o

f aff

ecte

d SS

Cs, Δ

CD

F AV

E, Δ

LER

F AV

E

In a

dditi

on to

the

disc

ussi

on p

rovi

ded

in It

em 4

.3.1

, sen

sitiv

ity st

udie

s are

requ

ired

to a

ssur

e ro

bust

dec

isio

n m

akin

g. C

hang

e in

risk

met

rics a

re u

sed

to d

eter

min

e th

e ris

k si

gnifi

canc

e an

d ris

k ac

cept

abili

ty o

f the

pro

pose

d ch

ange

.

An

exam

ple

is e

valu

atin

g th

e ch

ange

in ri

sk a

ssoc

iate

d w

ith re

laxa

tion

of Q

A re

quire

men

ts.

5. O

VE

RSI

GH

T A

CT

IVIT

IES

5.1

Per

form

ance

mon

itori

ng

5.1.

1 P

lann

ing

and

prio

ritiz

atio

n of

insp

ectio

n ac

tiviti

es (r

egul

ator

y an

d in

dust

ry)

PSA

bas

ed ra

nkin

g of

des

ign

and

oper

atio

nal f

eatu

res i

s us

ed to

focu

s res

ourc

es fo

r reg

ulat

ory

and

indu

stry

in

spec

tions

on

impo

rtant

issu

es a

nd e

quip

men

t.

CD

F AV

E, LE

RF A

VE,

QH

Os,

risk

impo

rtanc

e m

easu

res o

f all

SSC

s and

HEs

, pr

imar

y co

ntrib

utor

s to

risk

, ΔC

DF A

VE,

ΔLE

RF A

VE,

CC

DP,

C

LER

P

The

insi

ghts

from

bas

elin

e PS

A re

sults

are

use

d to

supp

ort s

ettin

g th

e ag

enda

and

prio

ritie

s for

sp

ecifi

c in

spec

tions

so th

at th

e ar

eas o

f the

pla

nt a

nd o

pera

tor a

ctio

ns th

at a

re m

ost r

isk

sign

ifica

nt

refle

ct th

e hi

ghes

t prio

rity.

An

exam

ple

coul

d be

the

deci

sion

to fo

cus t

he sc

ope

of a

n in

spec

tion

on th

e m

ater

ial c

ondi

tion

of

SSC

s fou

nd to

be

resp

onsi

ble

for t

he d

omin

ant r

isk

sequ

ence

s in

the

plan

t’s P

SA.

156

Brie

f des

crip

tion

of P

SA a

pplic

atio

n PS

A re

sults

and

m

etric

s for

use

in

deci

sion

mak

ing9

Com

men

ts o

n ho

w P

SA m

odel

s can

be

used

to su

ppor

t app

licat

ion

and

exam

ples

5.1.

2 L

ong

term

ris

k-ba

sed

perf

orm

ance

indi

cato

rs

The

long

term

risk

bas

ed in

dica

tors

focu

s on

mon

itorin

g pl

ant b

ehav

iour

in o

rder

to g

et in

sigh

ts o

n th

e pa

st h

isto

ry

of N

PP sa

fety

and

to u

pdat

e th

e ca

lcul

ated

ave

rage

CD

F.

Long

term

use

incl

udes

ana

lysi

s of p

ast p

lant

beh

avio

ur

inte

grat

ing

the

even

ts o

ccur

red,

failu

res a

nd

unav

aila

bilit

ies.

This

info

rmat

ion

(incl

udin

g C

DF

trend

s, co

mpa

rison

bet

wee

n ex

pect

ed a

nd c

alcu

late

d C

DF,

etc

.) is

of

inte

rest

to re

gula

tors

and

hig

h-le

vel p

lant

man

agem

ent.

Long

term

risk

bas

ed in

dica

tors

can

als

o he

lp to

pin

poin

t ag

ing

effe

cts o

n co

mpo

nent

s and

syst

ems.

This

info

rmat

ion

is im

porta

nt fo

r the

pla

nt st

aff a

nd c

an in

itiat

e de

sign

ch

ange

s or m

odifi

catio

ns to

test

ing

and

mai

nten

ance

st

rate

gies

, etc

. Sim

ilarly

, lon

g te

rm ri

sk in

dica

tors

can

be

draw

n up

for p

lann

ing

purp

oses

. For

long

term

pla

nnin

g,

the

assu

mpt

ions

rega

rdin

g pl

anne

d de

sign

cha

nges

, ex

pect

ed c

ompo

nent

beh

avio

ur, e

tc. c

an b

e in

trodu

ced

in

the

PSA

mod

els a

nd d

ata

and

can

be a

naly

sed

to o

btai

n th

e ex

pect

ed a

vera

ge C

DF

for t

he n

ext p

erio

d.

CD

F AV

E, LE

RF A

VE,

CD

F{t}

, LER

F{t}

, Q

HO

s, ris

k im

porta

nce

mea

sure

s of a

ll SS

Cs a

nd H

Es,

prim

ary

cont

ribut

ors

to ri

sk

The

PSA

resu

lts c

an b

e us

ed to

det

erm

ine

the

appr

opria

te se

t of p

erfo

rman

ce in

dica

tors

. Fo

r ex

ampl

e, th

e hi

gh ri

sk si

gnifi

cant

SSC

s can

be

used

to d

efin

e SS

C u

nava

ilabi

lity

and

failu

re

perf

orm

ance

met

rics t

hat a

re h

ighl

y co

rrel

ated

to si

gnifi

cant

CD

F an

d LE

RF

impa

cts.

If th

ese

risk

met

rics a

re u

pdat

ed o

ver a

long

per

iod

of ti

me,

agi

ng e

ffec

ts m

ay b

e in

dica

ted.

Spec

ial i

ndic

ator

s mig

ht b

e us

eful

that

are

der

ived

from

pla

nt sp

ecifi

c da

ta a

nd o

pera

ting

expe

rienc

e. F

or th

is p

urpo

se, t

he u

se o

f pla

nt-s

peci

fic c

ompo

nent

relia

bilit

y da

ta is

impo

rtant

. The

co

mpo

nent

s and

SSC

to b

e an

alys

ed c

an b

e de

rived

with

the

use

of im

porta

nce

mea

sure

s.

Ris

k M

onito

r can

be

used

as a

supp

ortin

g to

ol to

der

ive

aver

aged

CD

F es

timat

es d

eriv

ed b

y in

tegr

atio

n of

inst

anta

neou

s CD

F fo

r the

obs

erve

d pl

ant c

onfig

urat

ions

.

An

exam

ple

is th

e tre

ndin

g of

the

num

ber o

f fai

lure

s or u

nava

ilabl

e ho

urs o

f a h

ighl

y ris

k si

gnifi

cant

SSC

s.

5.1.

3 S

hort

term

ris

k ba

sed

perf

orm

ance

indi

cato

rs

Ris

k ba

sed

indi

cato

rs fo

r sho

rt te

rm u

se re

quire

in

stan

tane

ous e

valu

atio

n of

risk

. Thi

s typ

e of

app

licat

ion

prov

ides

info

rmat

ion

on c

hang

es in

CD

F du

e to

pla

nt

even

ts a

nd ri

sk a

ssoc

iate

d w

ith p

lann

ed a

ctiv

ities

.

CD

F AV

E, LE

RF A

VE,

CD

F{t}

, LER

F{t}

, Q

HO

s, ris

k im

porta

nce

mea

sure

s of a

ll SS

Cs a

nd H

Es,

prim

ary

cont

ribut

ors

to ri

sk

The

PSA

resu

lts c

an b

e us

ed to

det

erm

ine

the

appr

opria

te se

t of p

erfo

rman

ce in

dica

tors

. Fo

r ex

ampl

e, th

e hi

gh ri

sk si

gnifi

cant

SSC

s can

be

used

to d

efin

e SS

C u

nava

ilabi

lity

met

rics t

hat a

re

high

ly c

orre

late

d to

sign

ifica

nt C

DF

and

LER

F im

pact

s.

Sim

ilar t

o Ite

m 5

.1.2

, use

of p

lant

-spe

cific

dat

a is

impo

rtant

.

Ris

k M

onito

r is a

n ap

prop

riate

tool

to e

valu

ate

inst

anta

neou

s CD

F.

An

exam

ple

is th

e id

entif

icat

ion

of th

e ab

norm

al c

ondi

tions

for e

quip

men

t ope

ratio

n le

adin

g to

in

crea

se in

thei

r fai

lure

pro

babi

litie

s.

157

Brie

f des

crip

tion

of P

SA a

pplic

atio

n PS

A re

sults

and

m

etric

s for

use

in

deci

sion

mak

ing9

Com

men

ts o

n ho

w P

SA m

odel

s can

be

used

to su

ppor

t app

licat

ion

and

exam

ples

5.2

Perf

orm

ance

ass

essm

ent

5.2.

1 A

sses

smen

t of i

nspe

ctio

n fin

ding

s

This

type

of a

pplic

atio

n pr

ovid

es in

form

atio

n on

cha

nges

in

risk

mea

sure

s ass

ocia

ted

with

insp

ectio

n fin

ding

s. C

hang

e in

risk

met

rics a

nd c

ondi

tiona

l ris

k m

etric

s can

be

used

to e

valu

ate

the

risk

impa

ct o

f deg

rada

tions

or i

ssue

s th

at a

re fo

und

durin

g th

e in

spec

tions

and

to e

valu

ate

poss

ible

cor

rect

ive

actio

ns.

Ris

k im

porta

nce

mea

sure

s of a

ll SS

Cs a

nd H

Es,

prim

ary

cont

ribut

ors

to ri

sk, C

DF A

VE,

LER

F AV

E, Δ

CD

F AV

E, Δ

LER

F AV

E

This

app

licat

ion

is si

mila

r to

the

risk

eval

uatio

n of

sign

ifica

nt e

vent

s in

term

s of h

ow th

e PS

A is

us

ed to

eva

luat

e th

e ris

k si

gnifi

canc

e of

a p

lant

con

ditio

n th

at m

ay b

e co

nsid

ered

for d

iffer

ent

leve

ls o

f ins

pect

ions

dep

endi

ng o

n th

e re

sults

. O

ften,

sim

plifi

ed g

ener

ic P

SA m

odel

s are

use

d to

pe

rfor

m a

con

serv

ativ

e sc

reen

ing

eval

uatio

n fir

st a

nd if

sign

ifica

nt, i

t may

be

follo

wed

up

with

a

mor

e re

alis

tic a

nd d

etai

led

eval

uatio

n. D

epen

ding

on

the

area

of i

nspe

ctio

n fin

ding

s, PS

A o

f di

ffere

nt sc

ope

mig

ht b

e ne

eded

.

Exam

ple:

Use

of P

SA m

odel

s to

estim

ate

the

risk

sign

ifica

nce

of a

n in

spec

tion

findi

ng th

at a

fire

ba

rrie

r had

bee

n im

prop

erly

rem

oved

from

a N

PP.

5.2.

2 E

valu

atio

n an

d ra

ting

of o

pera

tiona

l eve

nts

By

PSA

bas

ed e

xtra

pola

tion

of o

pera

tiona

l eve

nts t

o ac

cide

nt sc

enar

ios w

ith se

rious

con

sequ

ence

s, va

luab

le

insi

ghts

can

be

gain

ed re

gard

ing

acci

dent

s on

the

basi

s of

min

or in

cide

nts,

with

out s

uffe

ring

thei

r rea

l con

sequ

ence

s. PS

A c

an b

e us

ed to

ana

lyse

pla

nt e

vent

s, w

hich

may

in

itiat

e a

plan

t trip

, deg

rade

or d

isab

le sa

fety

syst

ems,

or

both

sim

ulta

neou

sly.

The

app

licat

ion

can

then

pro

vide

an

estim

ate,

in te

rms o

f a c

ondi

tiona

l pro

babi

lity,

of t

he

avai

labl

e m

argi

n fo

r an

acci

dent

with

una

ccep

tabl

e co

nseq

uenc

es. T

hus,

the

basi

c pu

rpos

e of

PSA

bas

ed

oper

atio

nal e

vent

ana

lysi

s is t

o de

term

ine

how

an

oper

atio

nal e

vent

cou

ld h

ave

dege

nera

ted

into

an

acci

dent

w

ith m

ore

serio

us c

onse

quen

ces a

nd to

der

ive

the

cond

ition

al p

roba

bilit

y of

cor

e da

mag

e du

e to

such

eve

nt.

CC

DP,

CLE

RP,

C

DF A

VE,

LER

F AV

E If

the

even

t in

ques

tion

is a

n in

itiat

ing

even

t, th

e PS

A m

odel

is u

sed

to e

stim

ate

the

CC

DP

and

CLE

RP,

who

se v

alue

s are

use

d to

det

erm

ine

the

safe

ty c

lass

ifica

tion

of th

e ev

ent.

The

pre

curs

or

even

t ana

lysi

s is a

lso

part

of th

is a

pplic

atio

n.

If th

e ev

ent i

n qu

estio

n im

pact

s the

ava

ilabi

lity

of o

ne o

r mor

e SS

Cs a

nd/o

r ope

rato

r act

ions

but

is

not a

n in

itiat

ing

even

t, th

e PS

A m

odel

is u

sed

to c

alcu

late

the

CD

F an

d LE

RF

taki

ng in

to a

ccou

nt

the

unav

aila

bilit

y of

the

affe

cted

SSC

s. R

isk

Mon

itor i

s an

appr

opria

te to

ol to

eva

luat

e th

e im

pact

of

such

eve

nts.

The

PSA

mod

el m

ust b

e ca

pabl

e of

eva

luat

ing

the

appr

opria

te im

pact

s ass

esse

d fo

r the

eve

nt.

Exam

ple:

Eva

luat

ing

the

cond

ition

al p

roba

bilit

y of

cor

e da

mag

e or

larg

e ea

rly re

leas

e fr

om a

si

gnifi

cant

safe

ty e

vent

such

as a

n in

itiat

ing

even

t acc

ompa

nied

by

degr

adat

ion

or fa

ilure

of

mul

tiple

SSC

s and

/or h

uman

act

ions

.

6. E

VA

LU

AT

ION

OF

SAFE

TY

ISSU

ES

6.1

Ris

k ev

alua

tion

6.1.

1 R

isk

eval

uatio

n of

cor

rect

ive

mea

sure

s

Bas

ed o

n PS

A in

sigh

ts c

orre

ctiv

e m

easu

res r

egar

ding

sa

fety

issu

es a

re d

evel

oped

. Thi

s may

incl

ude

expl

orat

ory

inve

stig

atio

n on

diff

eren

t var

iant

s to

reso

lve

a pa

rticu

lar

issu

e.

ΔC

DF A

VE,

ΔLE

RF A

VE

Cha

nge

in ri

sk m

etric

s are

use

d to

det

erm

ine

the

risk

sign

ifica

nce

and

risk

acce

ptab

ility

of t

he

prop

osed

cha

nge

base

d on

risk

cha

ract

eriz

atio

n.

Exam

ple:

Ris

k ev

alua

tion

of m

easu

res t

aken

to re

duce

the

risk

of re

acto

r ves

sel h

ead

corr

osio

n.

158

Brie

f des

crip

tion

of P

SA a

pplic

atio

n PS

A re

sults

and

m

etric

s for

use

in

deci

sion

mak

ing9

Com

men

ts o

n ho

w P

SA m

odel

s can

be

used

to su

ppor

t app

licat

ion

and

exam

ples

6.1.

2 R

isk

eval

uatio

n to

iden

tify

and

rank

safe

ty is

sues

As a

resu

lt of

a P

SA, i

mpo

rtant

new

pla

nt sp

ecifi

c sa

fety

is

sues

and

gen

eric

issu

es m

ay b

e id

entif

ied.

Fur

ther

mor

e,

PSA

is u

sed

for e

valu

atin

g th

e re

lativ

e im

porta

nce

of

exis

ting

and

new

safe

ty is

sues

.

CD

F AV

E, LE

RF A

VE,

QH

Os,

risk

impo

rtanc

e m

easu

res o

f all

SSC

s and

HEs

, pr

imar

y co

ntrib

utor

s to

risk

, key

as

sum

ptio

ns

impa

ctin

g re

sults

Con

tribu

tors

to ri

sk a

nd ri

sk im

porta

nce

mea

sure

s are

use

d to

iden

tify

and

rank

safe

ty is

sues

. A

lso

safe

ty is

sues

iden

tifie

d ou

tsid

e th

e PS

A c

an b

e ev

alua

ted

by th

e PS

A to

det

erm

ine

thei

r ris

k si

gnifi

canc

e on

ce th

e is

sues

hav

e be

en a

sses

sed

for r

isk

char

acte

rizat

ion,

i.e.

det

erm

inat

ion

of

affe

cted

initi

atin

g ev

ents

, acc

iden

t seq

uenc

es, S

SCs a

nd o

pera

tor a

ctio

ns. S

ome

safe

ty is

sues

may

re

quire

ext

ensi

ons t

o PS

A m

odel

to e

valu

ate.

Exam

ple:

Elim

inat

ion

of a

n ite

m fr

om th

e lis

t of u

nres

olve

d sa

fety

issu

es b

ased

on

risk

insi

ghts

.

6.2

Reg

ulat

ory

deci

sion

s

6.2.

1 L

ong

term

reg

ulat

ory

deci

sion

s

PSA

insi

ghts

are

use

d to

gui

de lo

ng te

rm p

riorit

izat

ion

of

regu

lato

ry o

bjec

tives

and

requ

irem

ents

, and

of r

elat

ed

safe

ty re

sear

ch.

CD

F AV

E, LE

RF A

VE,

QH

Os,

risk

impo

rtanc

e m

easu

res o

f all

SSC

s and

HEs

, pr

imar

y co

ntrib

utor

s to

risk

, ΔC

DF A

VE,

ΔLE

RF A

VE,

CC

DP,

C

LER

P

PSA

resu

lts a

re u

sed

to d

evel

op ri

sk in

sigh

ts, a

nd st

rate

gies

to m

aint

ain

or re

duce

risk

leve

ls a

re

devi

sed.

Cha

nge

in ri

sk m

etric

s are

use

d to

eva

luat

e po

ssib

le c

hang

es to

requ

irem

ents

nee

ded

to

impl

emen

t the

risk

man

agem

ent s

trate

gy.

Exam

ple:

Dec

isio

n to

shut

dow

n th

e pl

ant o

r ter

min

ate

plan

t ope

ratio

n un

til th

e ne

cess

ary

glob

al

mod

ifica

tions

aim

ed to

redu

ce ri

sk a

ssoc

iate

d w

ith p

lant

ope

ratio

n w

ould

be

perf

orm

ed.

6.2.

2 In

teri

m r

egul

ator

y de

cisi

ons

PSA

is u

sed

to a

llevi

ate

a re

gula

tory

con

cern

, whi

le lo

nger

-te

rm so

lutio

ns c

an b

e ev

alua

ted.

Issu

es th

at ty

pica

lly

requ

ire a

n in

terim

dec

isio

n ar

e: (a

) nee

d fo

r reg

ulat

ory

actio

n in

resp

onse

to a

n ev

ent a

t a p

lant

, (b)

one

-tim

e ex

empt

ions

from

TS

or o

ther

lice

nsin

g re

quire

men

ts, a

nd

(c) t

empo

rary

mod

ifica

tions

to h

ardw

are

conf

igur

atio

n or

pr

oced

ures

.

CD

F AV

E, LE

RF A

VE,

QH

Os,

risk

impo

rtanc

e m

easu

res o

f all

SSC

s and

HEs

, pr

imar

y co

ntrib

utor

s to

risk

, ΔC

DF A

VE,

ΔLE

RF A

VE,

CC

DP,

C

LER

P

The

use

of P

SA in

rega

rd to

this

app

licat

ion

is e

ssen

tially

the

sam

e as

in It

em 6

.2.1

; dep

endi

ng o

n th

e su

bjec

t of t

he in

terim

regu

lato

ry d

ecis

ion

may

be

deal

ing

with

diff

eren

t ris

k ev

alua

tion

aspe

cts

(see

App

licat

ion

Gro

up 6

.1).

Exam

ple:

Dec

isio

n to

term

inat

e pl

ant o

pera

tion

until

the

mod

ifica

tions

aim

ed to

redu

ce ri

sk

asso

ciat

ed w

ith p

lant

ope

ratio

n w

ould

be

perf

orm

ed o

r dec

isio

n to

allo

w c

erta

in c

hang

es a

t the

pl

ant a

imed

to in

crea

se c

ost e

ffic

ienc

y of

pla

nt o

pera

tion

if in

sign

ifica

nt ri

sk in

crea

se w

ould

be

just

ified

.

159

Appendix III

MAPPING THE PSA ELEMENTS TO THE PSA TASKS ADDRESSED IN THE IAEA PSA GUIDELINES

The table below shows the relation between the PSA elements defined in this publication and PSA technical tasks from the IAEA PSA Procedure Guide ‘Procedures for Conducting PSA of NPPs (Level 1)’, Safety Series No. 50-P-4, 1992. The table helps to realize what tasks from the Procedure Guide are correspondent to the PSA elements.

ID PSA ELEMENT IN TECDOC TASKS FROM SAFETY SERIES No. 50-P-4

IE Initiating Events Analysis Task 13: Selection of initiating events

Task 17: Grouping of the initiating events

Task 24: Assessment of the frequency of initiating events

AS Accident Sequence Analysis Task 12: Definition of core damage states or other consequences

Task 14: Determination of safety functions

Task 15: Assessment of function/system relationships

Task 18: Event sequence modelling

Task 22: Impact of physical processes on development of logic models

Task 23: Classification of accident sequences into plant damage states

Task 27: Determination of accident sequence Boolean equations

SC Success Criteria and Supporting Analysis

Task 16: Assessment of plant system requirements (success criteria)

SY Systems Analysis Task 19: System modelling

Task 22: Impact of physical processes on development of logic models

HR Human Reliability Analysis Task 20: Human performance analysis

Task 26: Assessment of human error probabilities

DA Data Analysis Task 25a: Assessment of component reliability

Task 25b: Assessment of common cause failure probabilities

DF Dependent Failures Analysis Task 21: Qualitative dependence analysis

MQ Model Integration and CDF Quantification

Task 28: Initial quantification of the accident sequences

Task 29: Final quantification of the accident sequences

Task 30: Uncertainty analysis

RI Results Analysis and Interpretation Task 31: Importance and sensitivity analysis

Task 34: Preparation of documentation

161

REFERENCES

[1] INTERNATIONAL ATOMIC ENERGY AGENCY, Safety Assessment and Verification for Nuclear Power Plants, Safety Guide, IAEA Safety Standards Series No. NS-G-1.2, IAEA, Vienna (2002).

[2] INTERNATIONAL ATOMIC ENERGY AGENCY, Procedures for Conducting Probabilistic Safety Assessments of Nuclear Power Plants (Level 1), Safety Series No. 50-P-4, IAEA, Vienna (1992).

[3] INTERNATIONAL ATOMIC ENERGY AGENCY, Procedures for Conducting Probabilistic Safety Assessments of Nuclear Power Plants (Level 2): Accident Progression, Containment Analysis and Estimation of Accident Source Terms: A Safety Practice, Safety Series No. 50-P-8, IAEA, Vienna (1995).

[4] INTERNATIONAL ATOMIC ENERGY AGENCY, Procedures for Conducting Probabilistic Safety Assessments of Nuclear Power Plants (Level 3): Off-Site Consequences and Estimation of Risks to the Public: A Safety Practice, Safety Series No. 50-P-12, IAEA, Vienna (1996).

[5] INTERNATIONAL ATOMIC ENERGY AGENCY, Procedures for Conducting Common Cause Failure Analysis in Probabilistic Safety Assessment, IAEA-TECDOC-648, IAEA, Vienna (1992).

[6] INTERNATIONAL ATOMIC ENERGY AGENCY, Human Reliability Analysis in Probabilistic Safety Assessment for Nuclear Power Plants: A Safety Practice, Safety Series No. 50-P-10, IAEA, Vienna (1996).

[7] INTERNATIONAL ATOMIC ENERGY AGENCY, Applications of Probabilistic Safety Assessment (PSA) for Nuclear Power Plants, IAEA-TECDOC-1200, IAEA, Vienna (2001).

[8] THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS, Standard for Probabilistic Risk Assessment for Nuclear Power Plant Applications, ASME RA-S-2002, ASME, New York (2002).

[9] INTERNATIONAL ATOMIC ENERGY AGENCY, A Framework for a Quality Assurance Programme for PSA, IAEA-TECDOC-1101, IAEA, Vienna (1999).

[10] INTERNATIONAL ATOMIC ENERGY AGENCY, IPERS Guidelines for the International Peer Review Service. Second Edition. Procedures for Conducting Independent Peer Reviews of Probabilistic Safety Assessments, IAEA-TECDOC-832, IAEA, Vienna (1995).

[11] INTERNATIONAL ATOMIC ENERGY AGENCY, Defining Initiating Events for Purpose of Probabilistic Safety Assessment, IAEA-TECDOC-719, IAEA, Vienna (1993).

[12] THE AMERICAN NUCLEAR SOCIETY, External Events in PRA Methodology, American Nuclear Society Standard ANSI/ANS-58.21-2003, ANS (2003).

[13] US NUCLEAR REGULATORY COMMISSION, Guidelines on Modelling CCFs in PSA, NUREG/CR-5485 prepared by A. Mosleh, D. M. Rasmuson and F. M. Marshall for USNRC, USNRC, Washington, DC (1998).

[14] US NUCLEAR REGULATORY COMMISSION, Handbook of Human Reliability Analysis with Emphasis on Nuclear Power Plant Applications, NUREG/CR-1278, USNRC, Washington, DC (1983).

[15] US NUCLEAR REGULATORY COMMISSION, Technical Basis and Implementation Guidelines for a Technique for Human Event Analysis (ATHEANA), Revision 1, NUREG-1624, USNRC, Washington, DC (1999).

[16] INTERNATIONAL ATOMIC ENERGY AGENCY, Case Study on the Use of PSA Methods: Station Blackout Risk at Millstone Unit 3, IAEA-TECDOC-593, IAEA, Vienna (1991).

163

[17] US NUCLEAR REGULATORY COMMISSION, Severe Accident Risks: an Assessment for Five U.S. Nuclear Power Plants, NUREG-1150, USNRC, Washington, DC (Vol. 1 and Vol. 2: December 1990), (Vol. 3: January 1991).

[18] US NUCLEAR REGULATORY COMMISSION, Interim Reliability Evaluation Program Procedures Guide, NUREG/CR-2728, USNRC, Washington, DC (1983).

[19] INTERNATIONAL ATOMIC ENERGY AGENCY, Review of Probabilistic Safety Assessments by Regulatory Bodies, Safety Report Series No. 25, IAEA, Vienna (2002).

[20] INTERNATIONAL ATOMIC ENERGY AGENCY, Regulatory Review of Probabilistic Safety Assessment (PSA) – Level 1, IAEA-TECDOC-1135, IAEA, Vienna (2000).

164

ABBREVIATIONS

AC alternating current

AMP accident management procedure

AOT allowed outage time

AS accident sequence

ASME American Society of Mechanical Engineers

ATHEANA a technique for human event analysis

ATWS anticipated transient without scram

BDBA beyond design basis accident

BDD binary decision diagram

BWR boiling water reactor

CANDU CANada Deuterium Uranium

CCCG common cause component group

CCDP conditional core damage probability

CCF common cause failure

CCI common cause initiator

CDF core damage frequency

CLERP conditional large early release probability

DBA design basis accident

DC direct current

DG diesel generator

ECCS emergency core cooling system

EDG emergency diesel generator

EOP emergency operating procedure

EPZ emergency planning zone

ESF engineered safety feature

ET event tree

FMEA failure mode effect analysis

FT fault tree

165

F-V Fussell-Vesely (importance measure)

GA general attribute

HE human error

HEP human error probability

HFE human failure event

HPCI high pressure coolant injection

HPECC high pressure emergency core cooling

HRA human reliability analysis

I&C instrumentation and control

ICCDP incremental conditional core damage probability

ICLERP incremental conditional large early release probability

IE initiating event

IPSART international probabilistic safety assessment review team

ISI in-service inspection

ISLOCA interfacing system LOCA

IST in-service testing

LERF large early release frequency

LOCA loss of coolant accident

LOOP loss of offsite power

LWR light water reactor

MCP main coolant pump

MMI man-machine interface

MSIV main steam isolation valve

NDE non-destructive examination

NPP nuclear power plant

P&IDs piping and instrumentation diagram

PFM probabilistic fractural mechanics

PORV pressurizer power operated relief valve

PRA probabilistic risk assessment

166

PSA probabilistic safety assessment

PSF performance shaping factor

PWR pressurized water reactor

QA quality assurance

QHO quantitative health objective

RAW risk achievement worth

RHR residual heat removal

RI-ISI risk-informed in-service inspection

RPS reactor protection system

RPV reactor pressure vessel

SA special attribute

SAMG severe accident management guideline

SAR safety analysis report

SG steam generator

SGSV steam generator safety valve

SLOCA small LOCA

SSC systems, structures, and components

STI surveillance test interval

TECDOC technical document

THERP technique for human error rate prediction

TS technical specification

167

CONTRIBUTORS TO DRAFTING AND REVIEW

Alsop, C.J. British Energy/ NNC International Consulting (BENIC), United Kingdom

Alzbutas, R. Lithuanian Energy Institute, Lithuania

Angaloor, V.K. Safety Directorate NPCIL, India

Bagdonas, A. Ignalina NPP, Lithuania

Bennemo, L.J.G. Swedish Nuclear Power Inspectorate, Sweden

Berg, H.P. Bundesamt für Strahlenschutz, Germany

Bertucio, R. SCIENTECH, United States of America

Bradley, R.E. Nuclear Energy Institute, United States of America

Chakraborty, S. Swiss Federal Nuclear Safety Inspectorate, Switzerland

Comanescu, L. Atomic Energy of Canada Limited, Canada

De Gelder, P. AVN, Belgium

Dinnie, K.S. Nuclear Safety Solutions Ltd, Canada

Drouin, M. U. S. Nuclear Regulatory Commission, United States of America

El-Shanawany, M. International Atomic Energy Agency

Elter, J. Paks NPP, Hungary

Eriksson, P. S. RINGHALS, Sweden

Evans, M. G. K. Jacobsen Engineering Ltd, United Kingdom

Fleming, K. Karl N. Fleming Consulting Services, United States of America

Gabco, P. Bohunice NPP, Slovakia

Ghelbereu, S. Cernavoda NPP, Romania

Gheorghe, R. Canadian Nuclear Safety Commission (CNSC), Canada

Godinez, V. Comisión Nacional De Seguridad Nuclear Y Salvaguardias, Mexico

Gomez-Cobo, A. Nuclear Installations Inspectorate Health and Safety Executive, United Kingdom

169

Grantom, C.R. South Texas Project Nuclear Operating Company, United States of America

Grint, G. Nuclear Installations Inspectorate, United Kingdom

Gubler, R. Ingburo Dr. Reinhard Gubler, Switzerland

Habib, A. Pakistan Nuclear Regulatory Authority, Pakistan

Hahn, L. GRS, Germany

Hamar, K. Hungarian Nuclear Safety Directorate, Hungary

Hellstrom, P. RELCON AB, Sweden

Hellstrom, P. E. RELCON AB, Sweden

Hendrickx,I. Tractebel Engineering, Belgium

Hörtner, H. GRS, Germany

Husarcek, J. Slovak Nuclear Regulatory Authority, Slovakia

Hustak, S. Nuclear Research Institute Rez, Czech Republic

Ilieva, M. Risk Engineering Ltd, Bulgaria

Ishaq, S. Karachi Nuclear Power Complex, Pakistan

Islamov, R. International Nuclear Safety Center, Ministry for Atomic Energy of the Russian Federation, Russian Federation

Jakes, M. Nuclear Safety State Office for Nuclear Safety, Czech Republic

Jang, S.C. Korean Atomic Energy Research Institute, Republic of Korea

Jonsson, O.B.V. OKG, Sweden

Kajimoto, M. Japan Nuclear Energy Safety Organization (JNES), Japan

Kaufer, B. OECD/NEA

Kichev, E. CENS

Kirchsteiger, C. European Commission, JRC, Netherlands

Klevtsov, S. ETD, Ukraine

Klügel, J.U. Goesgen NPP, Switzerland

Kobilicova, M. CENS

170

Koeberlein, K. GRS, Germany

Kolesov, S. National Nuclear Energy Generating Co. ‘EnergoAtom’, Ministry of Fuel and Energy, Ukraine

Kompella, J.D. Safety Directorate NPCIL, India

Kouzmina, I. International Atomic Energy Agency

Krasnukha, S. South-Ukrainian NPP, Ukraine

Kubanyi, J. European Commission Joint Research Centre

Kulig, M.J. ENCONET, Austria

Lanore, J.M. IRSN, France

Lehner, J. Brookhaven National Laboratory, United States of America

Lopez, R.M. National Commission of Nuclear Safety and Safeguards, Mexico

Lyubarskiy, A. SEC NRS Gosatomnadzor RF, Russian Federation

Macsuga, G. Nuclear Safety Directorate Hungarian Atomic Energy Authority, Hungary

Marinova, B. Risk Engineering Ltd, Bulgaria

May, R. European Commission, JRC, Netherlands

Mlady, O. Temelin NPP, Czech Republic

Moir, G. R. British Energy, United Kingdom

Morozov, V. Atomenergoproject, Russian Federation

Niehaus, F. International Atomic Energy Agency

Novakova, H. Relko Ltd, Slovakia

Papazov, V. Kozloduy NPP, Bulgaria

Parry, G. US NRC, United States of America

Patrik, M. Nuclear Research Institute Rez, Czech Republic

Petri, M.C. Argonne National Laboratory, United States of America

Prochaska, J. VUJE Trnava Inc. Engineering, Slovakia

Ranguelova, V. International Atomic Energy Agency

171

Reinhart, F.M. US NRC, United States of America

Rybar, J. Nuclear Regulatory Authority, Slovakia

Schoen, G. HSK, Switzerland

Schultz, R. HSK, Switzerland

Serbanesku, D. Pebble Bed Module Reactor (PBMR) Ltd, South Africa

Shapiro, H. S. Atomic Energy of Canada Limited, Canada

Sheronov, Y. Rovno NPP, Ukraine

Shiversky, E. RDIPE, Russian Federation

Sholly, S. Institute of Risk Research University of Vienna, Austria

Sloane, B. Westinghouse Electric Co. LLC, United States of America

Suransky, L. Mochovce NPP, Slovakia

Svirmickas, S. State Nuclear Power Safety Inspectorate (VATESI), Lithuania

Tokmachev, G. Atomenergoproject, Russian Federation

Tong, J. Institute Of Nuclear Energy Technology, Tsinghua University, People’s Republic of China

Tudor, C. Cernavoda NPP, Romania

Utenkov, S. ROSENERGOATOM, Russian Federation

Van Graan, H. Pebble Bed Module Reactor (PBMR) Ltd, South Africa

Varde, P.V. Bhabha Atomic Research Centre, India

Vazquez, T. CSN, Spain

Versteeg, M.F. Ministry of Housing, Spatial Planning and the Environment, Netherlands

Vojnovic, D. Slovenian Nuclear Safety Administration, Ministry of the Environment, Slovenia

Yang, J.E. Korean Atomic Energy Research Institute, Republic of Korea

Yllera, J. International Atomic Energy Agency

172

Documents

Determining the quality of probabilistic safety assessment (PSA) for