ML20094L488

Aging Research Information Conference - Abstracts of Papers

ML20094L488
Person / Time
Issue date:	03/31/1992
From:	Beranek A NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES)
To:
References
NUREG-CP-0121, NUREG-CP-121, NUDOCS 9203250301
Download: ML20094L488 (87)

Text

_ ___ ___ _ _____ _________ ______ __

N Ulti.(iCI'-0121 Aging Research Information Conference-Abstracts of Papers d

g.

ny,yn

=

g I

i o

lleid at iloliday Inn Crowne I'laza Rockville, Maryland March 24-27,1992 Office of Nuclear Regulatory Research U.S. Nuclear Regulatory Commission

,f.

g A-

88*??uR86'"

CP-0121 R PDR

o NOTICE These proceedings have boon authored by a contractor of the United States Government. Nolther the United i

States Government not any agency thereof, or any of thelt employcos, makes any warranty, expressed or implied, or assumes any legal liability or responsibility for any third party's use, or the tosults of such uso, of any information, apparatus, product or process disclosed in these proceedings, or represents that its use by such third party would not infringo privately owned rights. Thu views expressed in those procoodings are not necessarily those of the U.S. Nuclear Regulatory Commission.

Available from Superir, endent of Documents U.S. Government Printing Office P.O. Box 37082 Wastungton D.C. 20013 7082 and National Technical information Service Springfield, VA 22161

N Ulti.(1/Cl -0121 Aging Research Information Conference-Abstracts of Papers licid at lloliday Inn Crowne Plaza Rockville, Maryland March 24-27,1992 Date Published: March 1992 Compiled by A. lieranek General Chairmen:

M. Vagins S. K. Aggarwal Omce of Nuclear Regulatory Research U.S. Nuclear Regulatory Commission llosted ley llrookhagen Natic = *.aborat ory pa***.y, j

AllSI1(Arl'

'lhis report contains abstracts of papers to be presented at the Aring itescaich inloimation Con-fetence held at the lloliday Inn Crimne I'lara in llockville, Maiyland, on March 21-27, lW2. 'lhis confetence is held to disseminate research tesults in the area of nuelcar p(mer plant aging (tom pro-gratus sponsored by the Olliec of Nuclear llegulatory itescatch. U.S. Nucicar llegulatoiy Conunis-sion. 'lhe conference will also provide an opportunity for engincess and scientists f rom around the world to eschange technical information aint discuss future inter national cooperation. 'lhe abstiaets appear in the order in which they will be presented at the conference, and they are piouped by techni-cal session. 'lhe full papers and the agenda for the conterence will be pubbshed as separate docu-inents.

iii i

CONTENTS AGING RESEARCH INFORMATION CONFERENCE March 24-27,1992 ABSTRACTS Y.a32 Abstract.............................

iii March 24,199?

TECHNICAL SESSION 1 Aging Program Overview and General Aging Issues NRC Plant Aging Research Program - Overview 1

R. Bosnak (NRC)

Nuclear Power Plant Connon Aging Terminology...

2 J. Vora (NRC), G. Sliter (EPRI)

Recordkeeping Technology To Support Aging Management.

3 J. Dukelow (PNL)

Treattnent of Aging in Routine Surveillance and Testing Practices - Examples.......................

4 M. Lintz (PNL), J. Vora (NRC)

TECHNICAL SESSION 2 Aging Effects on Risk Assessment Degradation Modeling - A Key to Understanding 6

Ef fects o f Aging and Maintenance.................

P. Samanta, D. Stock (BNL), W. Vesely (SAIC),

and J. Vora (NRC)

O Risk Evaluations of Aging W. Vesely (SAIC), G. Weidenhamer (NRC)

A Technique of including the Effect of Aging of Passive 10 Components in Probabilistic Risk Assessments...........

J. Phillips (INEL), G. Weidenhamer (NRC) 12 Validation Issues in Aging Risk Evaluations M. Hassan, P. Samanta (BNL), W. Vesely (SAIC)

Maintenance Practices To Manage Risk Associated with 13 Aging-Related Safety issues W. Enderlin, T. Vo (PNL) y

March 25,1992 TECHNICAL SESSION 3A Aging Management Reactor Coolant Pressure Boundary Components Page Environmentally Assisted Cracking and fatigue of Reactor Structural Materiais in LWR Environments.........

16 T. Kassner et al. (ANL)

Risk-Based Inspection for Management of Aging Dearadation 16 T. Vo, F. Simonen (PNL), J. Muscara (NRC)

Improved in-service Inspection Program for Managenent o f Degrada tion in Steam Generator Tubing.............

17 R. Kurtz (PNL), J. Muscara (NRC)

Aging Management of Major LWR Components.............

19 V. Shah, U. Sinha, A. Ware (INEL)

Lessons learned from Fatigue Failures in Majo r LWR Componen ts.......................

21 A. Ware, V. Shah (INEL)

TECHNICAL SESSION 3B Aging Management Electrical & Mechanical Components and Systems !

A Comprehensive Approach To Manage Aging i n Nucl e a r Se rv i ce Wa te r Sys tems.................

23 A. Johnson, Jr., D. Jarrell (PNL), J. Burns (NRC)

Assessment of Diagnostic Methods for Determining Degradation of Motor-Operated Valves..,......

25 H. Haynes (ORNL), W. Farmer (NRC)

Life Testing of a low Voltage Air Circuit Breaker 26 M. Subudhi (BNL), S. Aggarwal (NRC)

Aging and low-Flow Degradation of Auxiliary feedwater Pumps 28 M. Adams (Case Western Reserve U.)

Aging Evaluation of Nuclear Plant RTDs and Pressure Transmitters 30 H. Hashemian (Analysis and Measurement Services Corp.)

vi

March 25,1992 lECHillCAL SESSION 4 A Aging Management Reactor vessel Page Power Reactor Embrittlement Data Base (PR-EDB): Uses in Evaluating Radiation Embrittlement of Reactor Vessels 31 T. Kam et al. (ORNL)

Aging Impact on the Safety and Operability of Nuclear Reactor Pressure Vessels 33 W. Pennell (ORNL)

The Application of Probabilistic Fracture Analysis to Residual Life Evaluation of Embrittled Reactor Vessels......

35 T. Dickson (ORtil), T. Simonen (PNL)

Managing Irradiation Embrittlement in Agin g Re ac to r Pre s s ure Ve s sel s..................

36 W. Corwin (ORl4L)

TECHNICAL SESSION 48 Aging Management Electrical & Mechanical Components and Systems !!

Detecting and Mitigating Aging in Component Cooling Water Systems 38 R. Lofaro (BNL), S. Aggarwal (t4RC)

Snubbe r Aging Asses smen t.....................

40 D. Brown (Lake), D. Blahnik (Pf4L), J. Burns (fiRC)

Managing the Aging of BWR Control Rod Drive Systems 42 R. Greene (ORtil), W. Farmer (NRC)

The Role of Monitoring and Trending Applied to Diesel Generator Aging 44 K. Hoopingarner (PNL), J. Burns (NRC)

Aging Studies of Batteries and Transforrers in Class lE Power Systems 4S J. Edson, E. Roberts (INEL) vii

l l

March 26,1992 TECHNICAL SESSION SA Aging Management Mechanical Components and Systems Pa ge Operating Experiences and Degradation Detection for Auxiliary Feedwater Systems 47 D. Casada (ORNL)

Aging Assessment of Residual Heat Removal Systems in Boiling Water Reactors 49 R. Lofaro (BNL), S. Aggarwal (NRC)

Corrosion and Erosion Effects on Valve Friction and Operabili ty......

51 T. Hunt, H. Magleby (INEL), G. Weidenhamer (NRC)

Effect of Component Aging on PWR Control Rod 53 Drive Sp9tems E. Grove et al. (BNL), S. Aggarwal (NRC)

TECHNICAL SESSION SB Aging Management Electrical & Mechanical Components and Systems 111 Aging Assessment of Reactor Instrumentation and Protection System Components, Phase I 54 A. Gehl, E. Hagen (ORNL), W. Farmer (NRC)

Operating Experience Review of Failures of Power Operated Relief Velves and Block Valves in Nuclear Power Plants......

50 G. Murphy (ORNL)

Assessment of Diagnostic Methods for Sol enoi d-Ope ra ted Val ve s.....................

57 R. Kryter (ORNL), W. Farmer (NRC)

Shippingport Station Aging Mancgement lessons...........

59 R. Aan (PNL), J. Burns (NRC) viii l

I March 26,1992 TECHNICAL SESSION 6A Aging Management Elec crical & Mechanical Components and Systems IV 89" Effectiveness of Surveillance Methods for the Class 1E Power and Reactor Protection Systems 61 V. Sharma (INEL)

Effective Aging Management of Circuit Breakers and Relays 63 J. Gleason (Wyle Labs) understanding and Managing Effects of Battery Charger and Inverter Aging................

65 W. Gunther (BNL), S. Aggarwal (NRC)

Aging As se ssment o f Cable s....................

67 M. Jacobus (SNL)

Assessment of Diagnostic Methods for Determining Degradation of Check Valves 69 H. HayneT (ORNL), W. Farmer (NRC)

TECHNICAL SESSION 6B Aging Management Structures, Structural Components, and Cast Stainless Steel An Overview of the Structural Agi.ig Program 70 D. Naus (ORNL), G. Arndt (NRC)

Data Base on Structural Materials Aging Properties........

72 C. Oland (ORNL)

Reliability-Based Condition Assessment of Concrete Structures in Nuclear Power Plants 73 B. Ellingwood, Y. Mori (Johns Hopkins U.)

Prediction of Aging Degradation of Cast Stainless Steel Components in LWR Systems 75 O. Chopra-(ANL)

Effect of Aging on the Predicted claximum Load-Carrying Capacity of Circumferentially Cracked Cast Stainless Steel Pipe..

76 P. Krishnaswamy, P. Scott (Battelle)

Evaluation of Aging Degradation of Structural Components.....

77

0. Chopra, W. Shack (ANL) ix

NRC Plant Aging Research Program - Overview Robert J. Bosnak Division of Engineering Office of Nuclear Regulatory Research U.S. Nuclear Regulatory Commission ABSTRAC_T Aging degradation in operating nuclear power plants must be managed to prevent safety margins from eroding below the levels provided in plant design

bases.

" Aging" is universal in nature.

No industrial complex including nuclear power plants (NPP) should be considered immune from its effects, for NPP, aging is manageable if its symptoms are recognized and predicted, if it is monitored, and if appropriate steps are taken for timely mitigation of age-related degradation.

The Division of Engineering's (DE) Plant Aging Retsarch program is a multi-branch,' multi-disciplined, integrated approach to understanding and managing age-related degradation in safety-related components, systems, and structures (CSS) by the Materials Engineering, the Electrical and Mechanical Engineering, and the Seismic and Structual Engineering Branches.

The hardware oriented engineering research is concerned with degradation of: 1) the components of the primary system pressure boundary (with principal attention paid to the reactor pressure vessel), 2) safety-related electrical and. mechanical components and systems, and 3) the civil engineering structures and materials, including containment.

Emphasis on the reactor pressure vessel stems from its importance to plant safety.

It is a component which is not redundant and for which failure is not acceptable.

The pressure vessel integrity research program is the oldest of three areas of aging research

.providing a high quality information data base to understand and make regulatory decisions' involving fracture mechanics; fatigue-life, including

-initiation and propagation; material properties and flaw growth; irradiation effects; neutron dosimetry and surveillance data analysis;-non-destructive examination (NDE)-techniques;.and annealing issues including validation testing.. Programs in the electrical and mechanical enginee,ing area and the civil engineering area are similarly developing the needed technical information data bases and guidance.for understanding and managing aging of selected CSS in nuclear power plants of all ages and types.

The aging management process central to these efforts, as developed by the DE research program, consists of three key elements:

1) select and prioritize components, systems, and structures for which aging must be managed, 2) identify and understand the relevant aging mechanisms and their effect on. the. propertas or performance of the selected CSS, and 3) take appropriate action to manage degradation through effective inspection, surveillance. condition monitoring, trending, preventive and corrective maintenance, and mitigation-to prevent reduction in safety margins.

This paper provides an overview of the aging-related research programs sponsored by the Division of Engineering, Office of Nuclear Regulatory Research.

1

i l

i NUCLEAR POWER PLANT COMMON AGING TERMIN0 LOGY JIT YORA GEORGE SLITER U.S. NUCLEAR REGULATORY COMMISSION ELECTRIC POWER RESEARCli INSTilVTE ABSTRACT The U.S. NRC and the nuclear industry programs and activities related to understanding aging mechanisms, condition monitoring, f ailure and residual life evaluations, and maintenance have created the need for developing uniform terminology in many of these areas.

A technical committee with nine members from the utility industry and regulatory research established the scope of nuclear power plant common aging terminology and used a systematic technical and lexicographical approach in developing ccmmon definitions.

The definitions _ cover 93 terms related to degradation, life cycle, and aging management of systems, structures, and components.

Results have been issued by the Electric Power Research Institute (EPRI) for t-tal use and comment. This paper gives an overview of the effort ano solicits constructive fe'dback.

2

j RECORDKLEPING TECHNOLOGY TO SMEEDR1 AGlhD MANAGLBEH1 J. S. Dukelow Pacific Northwest Laboratory Pacific Northwest Laboratory (*1 has investigated the capability of current recordkeeping technology to support aging management.

This paper discusse. technical issues associated with potential enhancements of nuclear n

plant records systems from the perspective of the lessons learned about equipment aging degradation mechanist.s and associated surveillance and monitoring techniques during the U.S. Nuclear Regulatory Commission's Nuclear Plant Aging Research Program, it considers both the specific types of technical data needed to ensure continued safe operation and the use of new technology to upgrade record systems.

Specific topics discussed in this paper include:

~

equipment reliability data needed to support the assessment of the impact of aging on the continued operation of the plant operational history data to support the assessment of residual life of mechanical and structural components and piping tools for the analysis and trending of equipment reliability data and operational history data design and implementation of plant record systems that will provide comprehensive and usable engineering design basis for the plant proposed improvements in the data input process for the plant records system compcterization of plant records systems, including conversion of existing records into machine readable forms.

(a)

Pacific Northwest Laboratory is operated for the U.S. Depai'tment of Energy under Contract DE-AC06-76RLO 1830.

This work was conducted under NRC FIN B2865.

3

l i

TREATMENT OF AGING IN ROUTINE SURVEILLANCE AND TESTING PRACTTCES - EXAMPLES M. P. Lintz Pacific Northwest Laboratory J. 6. Vora V. S. Nuclear Regulatory Commission Routine surveillance and testing are performed on components and systems that are important to the safety of nuclear power plants.

These surveillances and teets ensure the operability and availahility of those components and systems.

The results of these activities can be used to detect, monitor, and trend age-related degradation.

Further, this information/ data could be useful to focus the surveillance and monitoring of those components and systems susceptible to significant age-related degradation, optimize maintenance effectiveness, and maintain the safety envelope as plants advance in age.

Routine surveillance and testing practices (RSTPs) can be effectively used to detect and manage aging.

For example, RSTPs serve as the major means of aging detection in dynamic plant equipment such as pumps, valves, breakers, and switches.

However, the effects of aging have not always been recognized explicitly in the RSTPs. More specific attention could be paid to 1) the types of aging mechanisms that are active on the components and systems that are important to safety, 2) the most effective methods of mitigation to counter the effects of the subject aging mechanisms, and 3) the RSTPs that can be used to detect, trend, and augment control of age-related degradation.

Pacific Northwest Laboratory (8) under the Nuclear Plant Aging Re-search (NPAR) Program is evaluating RSTPs for representative components from three perspectives:

1) the extent to which they address age-related degrada-tion; 2) their potential contribution to accelerated aging and service wear (e.g., fast starts of diesal generators, and the loads imposed on the auxilia-ry feedwater pump while testing in a pumping mode); and 3) their capabilities for detecting the symptoms of aging.

These evaluations are performed by

1) reviewing, from an aging pers

2) recommending improved RSTPs (pective, representative components, and and, if applicable, the associated bases) to account for aging.

Examples of evaluations for the emergency diesel generator system and the service water system follow.

Emeroency Diesel Generator System The emergency diesel generator (EDG) system provides backup electrical power needed by a nuclear power plant in the event of a loss of offsite ac power. The EDG components include the instrument and control, fuel oil, starting, cooling, lubricating oil, intake, exhaust, and generating systems, (a)

Pacific Northwest Laboratory is operated for the U.S. Department of Energy under Contract DE-AC06-76RL0 1830. Work is conducted under NRC FIN B2865.

4

and the engine. An NPAR study determined that these components are subject to degradation attributable to corrosion, vibration-induced f atigue, oxidation, themal stresses, shock, contamination, biofouling, manuf acturing and design errors, improper or excessive operation, improper or inadequate maintenance, and adverse environmental conditions. Many, if unmitigated, may enhance with time and age.

Each of the component systems of the EDG is subject to its own set of degradation mechanisms, and each mechanism causes varying degrees of degradation.

Some RSTPs are effutive for detecting aging mechanisms, some are neutral, and some promote degradation.

The evaluation concluded with recommendations to improve and supplement the RSTPs.

Service Water System The three types of service water systems (SWSs) perform vital safety functions in nuclear reactors and are the final link between the reactor and the ultimate heat sink, which is typically a sea, river, lake, or cooling pond. SWSs also provide cooling to safety-related equipment, such as EDGs and emergency core cooling systems.

Depending on its design, all or part of a SWS may be exposed to raw or relatively aggressive treated water. Therefore, SWS components are subject to many age-related degradation mechanisms.

Aging in SWSs is a complex subject and is only now being studied in sufficient depth to focus on effective methods of mitigating aging mechanisms.

An NPAR study has identified the age-related degradation mechanisms that are active in SWSs and where within the SWS these mechanisms are operative.

The study has also summarized a review of the RSTPs to evaluate their effective-ness in detecting age-related degradation mechanisms, and has developed recommendations to improve and supplement the RSTPs.

Generic Lecter (GL) 89-13. " Service Water System Problems Affecting Safety-Related Equipment,"

dated July 18, 1989, contains specific recommendations to address SWS aging.

However, there are critical areas within the SWS tt.at the GL does not address, i

such as the need to verify that the degradation rates of heat exchanger tubes, tubesheets, and waterboxes will remain above the minimum test limits through the next surveillance interval. A similar statement could be made for other components in the SWS.

The NPAR recommendations augment those of the GL to form a complete set of RSTPs that focus on the aging management of the SWS.

5

l Degradation Modeling A Key to Understanding Effects of Aging and Maintenance

• Pranab samanta, David Stock, William Vesely,t and J i tendra Vora

Department of Nuclear Energy Brookhaven National Laboratory Upton, New York 11973 Science Applications International Corporation T

• • U.S. Nuclear Regulatory Commission

SUMMARY

Component degradation modeling is the analysis of component degradations for the purpose of developing models of the degradation process and its implications. Degradation modeling can encompass many dif ferent areas, from the microscopic modeling of material degradation processes to macroscopic modeling of times of occurrences of degradations, in this paper. we present basic concepts, approaches, applications, and data needs of degradation modeling using times of occurrences of component degradations and failures.

We discuss degradation modeling from the viewpoint of understanding the effects of aging and the role of maintenance in mitigating the aging effects.

We argue that degradation modeling is a key to understanding the effects of aging and maintenance and should be the principal focus of aging analysis.

Since degradations generally occur before failures, detecting aging trends in degradations allows the aging effects to be corrected before they impact failures.

Furthermore, degradations generally occur more frequently than I

failures, providing a larger data base for analyzing aging effects.

In this paper, we discuss the basic concepts and matFematical development of a simple degradation model where the operational performance of a component is divided into three states - normal operating state, degraded state, and a failure state.

Traditional reliability analysis treats only two states of a component - normal operating state and failure state. Using Markovian approaches and renewal theory, we establish relations among the atates using rates of degradation and failure occurrences.

The relations are used to define the estimates of the effectiveness of maintenance in preventing degradations frcm becoming failures.

Specific applications of the modeling approaches are performed for " active" components.

Both standby and continuously operating components are analyzed.

All the applications demonstrate the usefulness and benefits of analyzing

• Work performed under the auspices of the U.S. Nuclear Regulatory Commission.

I 6

component data using degr adat ion modeling approaches, in case of residual heat removal (RllR) system pumps, a standby " active" component, degradation rate shows a " bathtub" curve where a distinct, in:reasing aging t rend is observed at later ages.

The ptuup failure rate does not show any increasing aging trend for the same period demonstrating early indication of aging trends through analyses of component degradat ions. For air compressors, a continuously operat ing component, both the degradation and f ailure rate show aging trends. The failure rate, which is significantly lower than the degradation rate in the first three years, increases significantly faster than the degradation rate and reaches the same value towards the end of the ten year period analyzed.

The effectiveness of maintenance in preventing age-related degradations f rom transforming to f ailures is also determined which is useful in understanding the effectiveness of existing maintenance activities.

Degradation modeling approaches can have broader applications in aging risn studies, in defining the effeetive maintenance practices, and in analyzing component reliability performance. Ext ensions of degradation modeling approaches to study reliability effects of different maintenance intervals, different maintenance durations, and dif ferent maintenance ef ficiencies also are discussed.

This extension will help define the rna intenance activities to mitigate a gi r.g effects, complement to the evaluation of the effectivenens of existing maintenance practices. We also discuss additional application areas and the data needs of the degradation modeling approaches.

7

RISK EVALUATIONS OF AGING W.E. Vesely Science Applications international Corporation G.ll. Weidenhamer U.S. Nuclear Regulatory Commission Aging of active and passive components can cause significant risk increases if the aging is not effectively managed. it is therefore important to evaluate the risk effects of aging in order to prioritize aging contributors and to evaluate the effectiveness of aging management programs. A methodology for risk evaluations of aging has teen developed as a part of the Nuclear Plant Aging Research (NPAR) Program being conducted by the Office of Researth of the U.S.

Nuclear Regulatory Commission. The methodology has been reponciin NUREG/CR 5510 (1) and in Reference (2).

The methodology which has been developed allows any current probabilistic risk assessment 3

(PRA) to be used to quantify and prioritize aging effects. The PRA does not need to be completely requantified to incorporate aging effects, but instead suitable risk importance coef0cients are extracted from the PRA and are combined with aging models for the individual components. To apply the methodology, component aging failure rate data are required, which can be based on engineering knowledge and generic data as well as plant specific data. Useful sensitivity studies can also be carried out by systematically varying the component aging rates to determine the degree of aging control of a test and maintenance program and to identify the most aging impacting contt.butors.

The risk evaluations of aging methodology has been peer reviewed and has been applied to support the regulatory analysis for license renewal rulemaking as described in NUREG 1362 (3).

In the demonstrations and applications which have been carried out, even when many components are aging,very few significantly impact the core damage frequency. Ilowever,if these relatively few, risk-important contributors are not identified and are not effectively maintained,large core damage frequency increases can occur. Furthermore, existing test and maintenance programs may not adequately focus on the risk-important contributors.

This paper describes highlights of a procedures guide which has been recently completed as a further step in transfemng the technology to the user. The procedures guide is presently under review and will be published shonly as a NUREG/CR. Specific steps in transfomaing a standard PRA to an age dependent PRA are described, including specific aging models which can be used. The discussions focus on a Level 1 PRA, or equivalently a probabilistic safety assessment (PSA), which evaluates system unavailabilities and core damage frequency. However, the same

,rocedures can be applied to Level 2 or 3 PRAs which evaluate radioactive releases or public acalth risks. Procedures for prioritizing aging comributors and for evaluating the risk-effectiveness of aging management programs are described. Demonstrations and insights are also given.

REFERENCES

1. NUREG/CR-5510 " Evaluations of Core Melt Frequency Effects Due to Component Aging and Maintenance", June 1990.

8 l

l

i

- 2, W.E. Vesely," incorporating Aging Effects into Probabilistic Risk Analysis" Relitthility Encineerine and Systemjafety, Vol. 32, No. 3,1991, pp. 315 337.

3, NUREG 1362 " Regulatory Analysis for Proposed Rule on Nuclear Power Plant License Renewal, July 1990.

l i

9.

... ~. _ _

A Technique Of Including The Effect Of Aging Of Passive Components In Probabilistic Risk Assessments

• Jerry H. Phillips Idaho National Engineering Laboratory Gerald H. Weidenhamer U.S. Nuclear Regulatory Commission The probabilistic risk assessments (PRAs) being developed at most nuclear power plants to calculate the risk of core damage generally focus on the possible failure of active components.

The possible failure of 3assive components is given little consideration. We are developing metiods for selecting risk-significant passive components and including them in PRAs. The methods provide effective ways to prioritize these passive components for inspection, and where inspection reveals aging damage, mitigation or repair can he employed to reduce the likelihood of component failure. We demonstrated a method by selecting a weld in the auxiliary feedwater (AFW) system. The selection of this component was based on expert judgement of the likelihood of failure and on an estimate of the consequence of component failure to plant safety.

We then modified and used the PRAISE computer code to perform a probabilistic structural analysis to calculate the probability that crack growth due to aging would cause the weld to fail.

The PRAISE code was modified to include the effects of material properties with age and changing stress cycles. The calculation included the effects of mechanical loads and thermal transients considered in the design and the effects of thermal cycling caused by a leaking check valve. We modified an existing PRA (NUREG-1150 plant) to include the possible failure of the AFW weld, and then we used the weld failure probability as input to the modified PRA to calculate the change in plant risk with time. The results showed that if the probability of component failure is high, the effect on plant risk is significant.

However, this particular calculation showed little change in plant risk for 48 years of service. The success of this demonstration shows that this method could be applied to nuclear power plants.

The demonstration showed the method is too involved for handling a large number of passive components and therefor simpler metheds are needed. A simpler method identified was to screen to limit the number of passive components and for the limited list use simpler methods to estimate the aging failure rate. Once the list of passive components is identified and the aging failure rates are estimated the PRA can be modified to incorporate them, a.

Work sponsored by the U. S. Nuclear Regulatory Commission, Office of Nuclear Regulatory Research, under DOE Contract No. DE-AC07-761001670; Dr. G.

H. Weidenhamer, Technical monitor 10 j

3

@uJ.I The screening can be done in several ways. A method identified was to select passive components based on consequence.

The consequence of passive component failure can be estimated using the PRA as a guide and a plant specific review.

Active components in the PRA can be used as surrogates for the passive components to determine the "importance" or contribution to the risk.

Another method is to use this list and further reduce it by eliminating those that do not have an identifiable aging mechanism.

Another is to include all passive components that could lead to system malfunction.

In the analysis we performed at the INEL the modified version of the PRAISE code was used to estimate the aging failure rate of one specific weld, but this code is too large to efficiently calculate the failure rate of a larp number of passive components in a nuclear power plant.

Therefore techniques were identified to estimate the failure frequency on a large number of components.

These techniques include data analysis of actual f ailures, expert solicitation, questionnaires, simple structural probabilistic models, as well as the large models such as PRAISE.

Data analysis is the most direct approach.

Statistical analysis and aging techniques developed by NPAR can be used to determine the failure probability parameters if a number of failures have been observed--but this data is usually not available on passive component failures.

Expert solicitation can be used to estimate the failure frequency and " wear out" portion of the bathtub curve.

This technique has been demonstrated by a committee sponsored by the NRC, called Tirgalex, and by Pacific Northwest Laboratories in their work on risk-based inspection, but this technique is only as good as the experts and information available to

them, in addition, where it may be easy for an expert to estimate stress or temperature difference it can be difficult for an expert to estimate a probability from this information.

Questionnaires have been developed and used to estimate the failure frequency of passive components especially during the design _ phase of a project.

The questionnaire lists the parameters for individual components such as types of materials, numbers and relative size of mechanical and thermal cycles, etc. so that a failure frequency estimation can be made.

The basis of the failure frequency estimation is numerous probabilistic structural analysis calculations on similar piping systems.

Simple probabilistic structural analysis models are the most exciting development.

These computer models allow simple user friendly input in a model that is developed from the larger structural analysis codes.

Many calculations can be performed in a short period of time to perform optimization studies, specify inspection frequencies and accuracies necessary to obtain a specific failure probability. Larger models such as PRAISE can be used on very large consequence passive components where more accuracy is needed.

Screened passive components cen be included in a PRA and standard risk analysis results can be used to identify " target level" aging failure rates that must be maintained to control risk.

11

Validation Issues In Aging. Risk Evaluations

• Mahbubul Hassan, Pranab Samanta, and William Veselyi Department of Nuclear Energy Brookhaven National Laboratory Uptori, New York 31973 I cience Applications International Corporation S

SUMMARY

~ As nuclear power plants grow older, the aging plant components may be increasingly susceptible to failure with consequent potential for increase in plant risk.

Evaluation of plant risk from aging is important not only from a safety perspective but to implement effective and efficient aging managemont programs-to assure reliable plant operation as well.

Significant efforts have been made in recent years - to develop methodologies to evaluate the risk impact-from aging of nuclear power plant components and systems.

The aging risk evaluation approaches which have been developed are based on extension of Probabilistic Risk Assessment (PRA) techniques and can be used to predict overall risk to plant from aging and/or to prioritize detailed contributors-to risk from aging.

In these-approaches, the aging effects are incorporated into existing

- plant' PRAs to' predict plant risk behavior with age.'

Different approaches may be used to incorporate-the= aging effects into the PRA and_a number of parametric models are available to describe component or system behavior with age.

The

- prioritization is' determined through an evt.luation of the risk contribution of individual components and systems. There are, however, uncertainties in these evaluationa arising primarily from,-(a) sps.rseness of available component aging data, (b) assumptions in component aging behavior modeling, and (c) assumptions

-in plant risk models and in risk quantification approaches. These variabilities

in modeling. assumptions -- and data can be significant and can affect-the applicability of aging risk evaluations.

- This paper summarizes the issues involved. in t w validation of aging risk

- evaluations for different applications. How these issues-are relevant in aging--

risk estimations, aging prioritizations, and/or other inferences are discussed.

The objective of the validation study-is to evaluate the sensitivity of the aging risk evaluations to modeling and data variabilities and to identify applications

which L are robust against - such variabilities.

As a first step sensitivity

-studies using a NUREG-1150 plant are currently underway to assess the impact-of model and data uncertainties on component prioritizations. The results of this study will help identify prioritization schemos which are robust and meaningful

% - mlications.

Recent developments from this work are discussed,

• Work.porformed under the auspices of the U.S. Nuclear Regulatory Commission.

12

MADTENANCE PRACllCES TQla!LAGlllSKJSSQGAlEQllE AGING-RELATEQ_SALL]Y ISMLS W.1. Enderlin and T. V. Vo Pacific Northwest Laboratory As part of the Nuclear Plant Aging Research (NPAR) Program, supported by theNglearRegulatoryCommission(NRC),PacificNorthwestLaboratory (PNL) has developed a risk-based method that may be applied to establish maintenance practices for operating nuclear power plants.

The approach uses existing results of probabilistic risk assessments (pRA) and related studies and relevant industry experience to provide guidance for managing risk associated with component aging. The developed method consists of three major steps: the first step develops a graded approach to maintenance, which identifies and prioritizes risk significant systems, structures, and com-ponents (SSCs) to focus maintenance ef fectiveness; the second step analyzes and characterizes system and component aging degradation; the third step identifies safety significant SSCs and their associated aging failure mechanisms useful to develop an effective aging maintenance program.

This paper provides suggestions that may be considered during the development of maintenance practices intended to focus on the management of risk associated with SSC aging.

The suggestions can be applied to a) develop a risk-based graded approach to maintenance, b) assess the impact of the aging process on safety significant SSCs, c) develop a maintenance program that recognizes the aging of safety significant SCCs as a unique risk factor, and d) incorporate a plan for managing this risk area.

A risk-based graded approach to maintenance involves selective and judicious assignment of resources to maintain facilities and equipment based on site-specific risk quantification.

Thus, it is essential that a site-specific maintenance plan ensures that the risk associated with aging of safety significant SSCs be included in this risk quantification.

Two primary issues that relate directly to potential risk are 1) priority of performing maintenance and 2) the depth and extent of procedures, training, qualification of personnel, procurement control, quality assurance, reporting, and documen-tation.

For nuclear applications, techniques normally employed in formulating a graded approach to maintenance for a specific facility are the PRA, de-scribed in NUREG/CR-2300; Reliability Centered Maintenance (RCH), a technique borrowed from the aviation industry; a judicious review of both plant-specific and industry-wide experience, as documented in LERs, bulletins, notices, the Nuclear Plant Reliability Data System (NPRDS), and in-plant maintenance, operations, and quality control records.

The recorded results of a plant's self-assessment are also useful in formulating a graded approach.

There are, however, advantages and limitations associated with each of these techniques; (a) Pacific Northwest Laboratory is operated for the U.S. Department of Energy under Contract DE-AC06-76RLO 1830.

Work is conducted under NRC FIN 82865.

13

l hence, all three should be considered when formulating a graded approach for a specific site.

Assessing the effect of aging on safety significant SSCs, identified as a result of site-specific risk quantification, requires a thorough understanding of aging mechanisms and the stressors imposed by these mechanisms, it must be understood that equipment items and component parts age and deteriorate over long outage periods as well as during operating periods, due to numerous aging mechanisms.

Additionally, many system hardwarc failures that occur during service are traceable to built-in (latent) manufacturing defects that acceler-ate specific aging mechanisms.

These defects may pass quality inspection during manufacturing but finally become evident during plant operation.

In NPAR program studies, PNL has identified twenty-seven component aging mecha-nisms. A glossary of these issues is included in NUREG/CR-5490, Vol 1.

Results from related PNL studies regarding system and component prioritization activities may also be applicable, in developing an effective maintenance program that minimizes risk resulting from aging of safety significant SSCs, it is useful to review the lessons learned by others who have been confronted with similar issues.

Where appropriate, these lessons learned can be used for developing maintenance plans.

Feedback mechanisms, which continually improve and refine the program, are vital if the maintenance program is to address changing needs, such as the degradation of plant equipment due to aging.

To establish a broader perspective for managing age-related degradation, PNL analyzed effective maintenance activities used by two commercial indus-tries and two military organizations to manage the aging of systems and components.

The four programs considered were:

a) the U.S. commercial airline industry; b) the U.S. Air Force B-52 bomber; c) the U.S. Navy Ballistic submarine; and d) the Japanese nuclear power industry.

The mainte-nance programs of these four organizations offered lessons potentially useful for managing aging in commercial nuclear power plants.

A summary of the maintenance-related activities to address system and component aging, based on the approach taken by each of these four organizations, is presented in Table 1 of the final report for this study (PNL-7823).

PNL also reviewed mainte-nance practices at various U.S. commercial nucicar power plants.

The suggestions set forth in this paper could be useful in establishing overhaul frequency, developing inservice inspection programs, developing outage plans and establishing control procedures.

These suggestions could also be useful in selecting the proper balance among maintenance methods (e.g., preventive maintenance versus corrective maintenance), selecting replacement components, especially where diminishing manufacturing source is an issue, and establishing design mod',fication needs to ensure adequate maintainability with respect to aging issues. Moreover, the approach set forth has the potential for enhancing maintenance effectiveness to manage aging by providing suggestions for focusing maintenance on risk significant SSCs.

14

i Environrnentally Assisted Cracking and Fatigue of Reactor Structural Materials in LWR Environtnents' T. F. Kassner, W. E. Ruther,11. M. Chung, J. Y. Park, P. D. lincks, D. R. Dierrks, and W. J. Shack Materials and Components Technology Division Argonne National Laboratory, Argonne, Illinois The fatigue life of A533-Gr B pressure vessel steel in high-purity (llP) deoxygenated water, in simulated PWR water, and in air was studied. The material for the tests was a medium-sulfur-content (0.016% S) steel obtained from the lower head of the Midland reactor. The fatigue data collected to date lie above the ASME design curve. Because fatigue crack growth rates (CGRs) in this class of material can show significant environmental enhancement in simulated PWR water under certain loading conditions, the effects of load shape and loading rate will be further explored in subsequent testing.

Fracture-mechanics CGR tests have been perfonned on composite specimens of A533-Gr B/inconel-182/Inconel-600 plated with nickel, and on homogeneous specimens of A533-Gr B material plated with chrome and nickel. Conventional unplated specimens have also been tested to provide baseline data. In the tests to date, plated specimens have been more susceptible to environmentally enhanced cracking than unplated specimens. The proposed revised ASME Section XI da/dN versus AK culves were used to calculate equiva-lent da/dt versus Kmax curves for comparison. Even for the plated specimens, the data are in reasonably good agreement with the predictions for all data from water containing 200 ppb dissolved oxygen.

Irradiated austenttic SSs can become susceptible to SCC. This susceptibility been attributed to radiation induced segregation (RIS) or depletion of elements such as Si, P S, Ni,and Cr. High (lip)- and commercial-purity (CP) Type 304 SS specimens were obtained from control blade absorber tubes after irradiation to fluences of up to 2 x 1021 n.cm-2 (E > 1 MeV) from two operating BWRs. Microchemical and microstructural changes in the steels were studied by Auger electron spectroscopy. Significant RIS of Si, P NI, and an unidentified element or compound associated with an Auger energy peak at 59 eV was observed in the CP material. Except for Ni, such segregation was negligible in the llP material.

No evidence of S segregation was observed in either material.

However, chromium depletion from grain boundaries was more pronounced in the HP than in the CP material. Slow-strain-rate-test were conducted on the CP and IIP materials in air and in simulated BWR water. The 11P material tuowed significant intergranular stress corrosion cracking in water and the strain to failure for com; arable fluence levels was lower in the llP material than in the CP material.

• Work supported by the omce of Nuclear Regulatory Research. U.S. Nuclear Regulatory Commission FIN A22122, Program Manager Dr. J. Muscara.

15 1

l RISK-BASED INSPECTION FOR MANAGEMENT OF AGING DEGRADATION T. V. Vo, F. A. Simonen, Pacific Northwest Laboratory J. Muscara, U.S. Nuclear Regulatory Commission Pacific Northwest Laboratory P. O. Box 999 Richland, WA 99352 As part of the Nondestructive Evaluation Reliability Program, sponsored by the Nuclear Regulatory Commission (NRC), Pacific Northwest Laboratory (PNL) is developing methods that use risk-based approaches to establish in-service inspection (ISI) plans of nuclear power plant components. The method first uses results of probabilistic risk assessment and failure modes and effects analysis technique to identify and prioritize the most risk-important systems and components for inspection priorities. Once high-priority components have oeen identified, an approach recommended by the American Society of Mechanical

~

Engineers (ASME) Research Task Force on Risk-Based Inspection Guidelines is used to determine the target (acceptable) risk and failure probability values for individual components.

Aging degradation that may impact on the estimated component failure probabilities are addressed.

inspection programs (method, frequency, extent) are then developed to manage these aging mechanisms.

Probabilistic structural mechanics techniques will be applied to establish inspection strategies that will ensure that component failure probabilities are maintained at acceptable level.

After candidate inspection strategies yielding component failure probabilities less than identified target values have been determined, decision analysis techniques can be used to optimize inspection strategies.

The Surry Nuclear Power Station, Unit I was selected for demonstrating the methodology.

The specific systems selected for analysis were the reactor pressure vessel, the reactor coolant, the low pressure injection, and the auxiliary feedwater. The results provide a risk-based ranking of components within these systems, and provides a basis for relating the proposed target risk to acceptable failure probability values for individual components.

Aging mechanisms (e.g., irradiation damage, embrittlement, erosion / corrosion, etc.) with the potential to impact on target rupture probabilities of components are also discussed.

Because of similarities in objectives, the PNL program is coordinated with an ASME Research Task Force on Risk-Based Inspection Guidelines. The initial task force document has made general recommendations on the application of risk-based methods to ISI, and forms the basis of future proposals to ASME for improved codes and standards.

Results of PNL studies are being made available to the ASME group to demonstrate the usefulness of the risk-based methodology.

16

Improved In-service Inspection Program for Management of Degradation in Steam Generator Tubing Richard Kurtz, Pacific Northwest Laboratory Joseph Muscara, U. S. Nuclear Regulatory Commission Pacific Northwest Laboratory P.O. Box 999 Richland, WA 99352 This pr.per discusses extensive research supported by the U. S. Nuclear Regulatory Conmission to develop information on the failure pressure of degraded steam generator tubing and the reliability and effectiveness of eddy current (ET) in service inspection (ISI) techniques to detect and size degradation, in 1976, the NRC authorized the Pacific Northwest Laboratory to conduct the Steam Generator Tube Integrity Program to develop the needed information.

A main objective of the program was to establish validated models, based on experimental data, for predicting the failure pressure of service-degraded tubing under normal operating and accident loading conditions.

More than 600 specimens of Inconel 600 tubing were mechanically flawed to simulate known or postulated defects present in steam generators.

These specimens were burst and collapse tested at steam generator operating tempcratures under controlled loading conditions.

From the data, constitutive aquations were developed relating failure pressure to flaw size and geometry.

The constitutive equations were validated by testing tube specimens with artificial flaws created by chemical means and by testing service-degraded tubes taken from the retired-from-service Surry 2A steam generator.

Information on the reliability of ET inspection techniques to detect and size flaws in laboratory and service-degraded tubes was developed throughout all phases of the program.

The most extensive and realistic data base was obtained from round robin examinations of the Surry 2A steam generator.

four round robins were performed to determine the reliability of conventional multi-frequency ET inspections and alternative NDE methods.

To validate the in situ NDE results, more than 550 tube segments were removed from the generator.

Pitting and wastage were the predominant tube defects found in the specimens examined.

Estimates of the probability of detection (POD) and sizing accuracy were obtained by matching the ET inspection results with data from both visual and destructive metallographic analysis of the removed specimens. Results indicated that the POD depended mainly on flaw severity.

For pitting / wastage type flaws, the P00 increased with wall loss and approached 0.9 for flaws greater than 40% through-wall.

Wide variations in the reported ET depth estimates were observed between specimens with similar wall-loss for an individual team and also within the same specimen for data from different inspection teams.

The team-to-team variations for a givea specimen appear to result from differences in analysis 17

1 procedures or the analyst's interpretation of the complex ET signal patterns.

Defect morphology and distribution within the corroded region was considered the major cause for the variations between specimens with similar wall-loss.

However, dents and surface deposits near the defects also contributed to the sizing variations, in general, El tended to undersize the pitting / wastage type defects, especially for the severely-degraded specimens.

Improved sizing accuracy was noted for one team that employed special frequency mixes to enhance the signal-to-noise ratio by suppression of sionals due to denting, copper deposits and support plates.

Also, ultrasonic and rotating ET probes were successfully used to augment conventional ET/ bobbin-coil data to obtain improved sizing accuracy.

The results of the ET reliability studies were used to develop models of POD and ET sizing error to provide a basis for evaluating and comparing various two-stage sampling plans for steam generator tube inspection.

For this study the goal of ISI was to identify all defective tubes which could fail by leak or burst during reactor operation should a main steam line break occur.

In this work a defective tube was defined as one with 2 75% through-wall degradation at the time of 151.

This definition was developed from burst test data which indicated that, on average, tubes with degradation about one inch in axial extent and 85% through-wall would fail under main steam line break loading conditions. A 10% flaw growth rate per operating cycle was assumed to arrive at the definition of a defective tube.

Analytical and Monte Carlo simulation studies were performed which indicated that a 40% systematic sequential sampling plan was almost as effective as 100%

inspection, assuming some clustering of degraded tubes. This sampling strategy relies on two key concepts to achieve a high level of effectiveness.

First, a relatively large, uniformly distributed initial sample is used to maximize the probability of finding isolated defective tubes, and second, detection of tube degradation of any level triggers second-stage inspection to aid in finding defective tubes which may be in close proximity.

In order for this sampling strategy to be offective high inspection reliability for detection and sizing of degradation is needed, even when degradation is < 75%

through-wall. As a consequence, performance demonstration qualification criteria have been developed to provide high inspection reliability by establishing appropriate thresholds on POD and flaw sizing performance.

The thresholds were selected because of the need for high reliability for identifying defective tubes and good reliability for finding degraded tubes.

The POD curve and flaw sizing requirements were also based, in part, on the levels of performance observed in the ET reliability studies.

18

AGING MANAGEMENT OF MAJOR LWR COMPONENTS' V. N. Shah, U. P. Sinha. A. G. Ware Idaho National Engineering Laboratory j

i The Aging Assessment and Mitigation Project has comprehensively evaluated degradation mechanisms af fecting the structural integrity of the major light water reactor (LWR) components and has identified several options for managing their aging. The three primary degradation mechanisms acting on the components are embrittlement, fatigue, and corrosion (including stress corrosion cracking).

This paper focuses on the management of stress corrosion cracking (SCC) mechanisms: primary water stress corrosion cracking (PWSCC) of pressure boundary components in PWRs, and intergranular stress corrosion cracking (IGSCC) and irradiation-assisted stress corrosion cracking (IASCC) in BWR vessels. Effective aging management of the SCC mechanisms includes evaluation of interactions between design, materials, stressors, and environment; identification and ranking of the susceptible sites; relieble inspection of damage; mitigation of damage, including modifications in water chemistry; and repair and replacement using corrosion-resistant materials.

Primary water stress corrosion cracking has causod both axial and circumferential through-wall cracks in Alloy 600 components constituting the PWR primary pressure boundary.

Two major concerns about PWSCL failures are an unisolable pressure boundary leak and corrosion of any carbon or low-alloy steel base metal exposed to leaking borated coolant.

PWSCC cracks have been found in tubes on both hot-and cold-leg sides and in tube plugs of recirculating steam generators.

Recently, these cracks have also been reported in pressurizer instrument nozzles and heater sleeves and in control rod drive mechanism nozzles, fabrication records of all Alloy 600 components in PWRs need to be reviewed to estimate residual stresses and to characterize microstructure so that the components can be ranked according to their susceptibility to PWSCC, Components with high residual stresses and no intergranular carbides have high susceptibility if operating temperatures are high, and such components need to be inspected for PWSCC damage.

In add; tion, methods need to be developed for ultrasonic examination for boric acid corrosion of base metal around the failed Alloy 600 nozzles. Alloy 690, a PWSCC resistant material, may be used for replacement of failed Alloy 600 components.

Intergranular stress corrosion cracking has caused through-wall cracking in BWR pressure vessel nozzle welds and is a potential degradation mechanism for vessel interior attachment welds. Two major concerns are a crack initiating in the weld propagating into the vessel base metal and a leak through the primary pressure boundary.

Welds most susceptible to IGSCC are those having Alloy 182 weld material; having high residual, applied, and thermal tensile stresses; and having i Work sponsored by the U.S.

Nuclear Regulatory Commission, Office of Nuclear Regulatory

Research, under DOE Contract No.

DE-AC07-76001570; Dr. G. H. Weidenhamer, Technical Monitor.

i 19

R a high electrochemical potential. In addition, the IGSCC susceptibility of welds increases as the BWR coolant conductivity increases.

The estimated residual tensile stresses-at the attachment welds not subject to postweld heat treatment are high (280 to 480 MPa), whereas the stresses at the welds subject to postwelJ heat treatment are low (140 to 280 MPa), The applied stresses are generally low except at the welds attaching the jet pump riser brace to the vessel, Thermal stresses are due to differential thermal expansion of the weld and base metal.

The electrochemical potential of the attachment welds in the upper region of the vessel is high because of the high oxidizing power of the BWR coolant leaving the

{

core, where it is subject to radiolysis, Specialized equipment and techniques are being developed for remote automated inspection of penetration welds, for example, inspection of the incore monitor housing-to lower head welds.

An SCC monitor may be employed to estimate any IGSCC crack initiation or growth in the nozzle and the attachment welds.

Recent SCC test results indicate that the hydrogen water chemistry, which has been found effective in suppressing 1GSCC in recirculation piping, has a potential to protect the attachment and nozzle weld materials. However, it does not provide the same level of protection to all weld locutions on the vessel, because the electrochemical potential varies, Susceptible welds in the upper region of the vessel, which have a higher electrochemical potential, will require a greater amount of hydrogen injection than the ones in the downcomer region or in the recirculation piping.

The greater amount of hydrogen injection has -

several adverse effects -that should be addressed, such as an intmse in the

-steam line radiation fields associated with an increased partit'ioning of N-16 in the steam phase, _ Damaged welds may be repaired by replacin: illoy L182 mMerial with the corrosion resistant Alloy 82 material, by applying - crrosion re.istant cladding to protect susceptible materials from exposure to WR coolut, and by using clad overlay as a short-term solution. Underwater wet welding and cutting techniques are being developed for the vessel repair.

Irradiation assisted stress corrosion cracking can occur e the nighly irradiated BWR reactor internal components fabricated from stain 4ss steel and nickel base alloys.

Several failures of the stainless steel compnents, such as an incore guide tube _and-a neutron monitor dry tube, are at'ribc ed to IASCC, Based on field experience and laboratory tests, IASCC det not accur below a certain threshold _ level of fast fluence; the threshold oecends on material, stress t

levels, geometries, and environment, ge thrpsho Nr hpe 304 stainless steel componergs witp high stresses i_s 5x10 n/cm:; -with lor stresses the threshold is 2x10 n/cm A relatively high level of dissolved oxygen and a crevice geometry - can lower the threshold and accelerate any existing IASCC damage -

Development of specia.lized equipment and - techniques for inspection, use of hydrogen water chemistry, and the.une vwater weld repair discussed above for the BWR vessel welds are also appi. cable here to manage IASCC-damage. Modified heat treatment and IASCC-resistant materials are being developed for replacement of damaged reactor internal comr-ts.

For example, preirradiation solution annealing treatments in the 12 J to 1300 C-temperature range can make Type 304 stainless steel _more resistant to IASCC.

High-purity Type 348 stainless steel is also found to be IASCC-resistant.

20

\\

~

LESSONS LEARNED FROM FATIGUE FAILURES IN MAJOR LWR COMPONENTS' A, G Ware, V. N. Shah Idaho National Engineering Laboratory Fatigue:is one of the leading degradation mechanisms affecting major light water reactor-(LWR) components, fatigue has caused surface cracks and, in some cases, through wall cracks and coolant leakage.

This paper evaluates the fatigue e

failure experience in the field and discusses the lessons learned that can be employed in managing fatigue damage. The lessons relate to stressors causing the fatigue damage; sites susceptible to the damage; and inspection, monitoring, and mitigation of the damage.

Investigation of field fatigue experience has identified stressors and degradation mechanisms that contributed to failures br*

e not accounted for in the designs.

Examples of such stressors are thermal stratification and

-striping acting on PWR surge lines, high-pressure safety injection lines, and feedwater nozzles. Design calculations for the fatigue usage factors for several affected components are being revised to account for these stressors.

Examples

-of the mechanisms, which were not. accounted for in the design, are environmentally assisted fatigue and high-cycle thermal fatigue. Environmentally assisted fatigue has caused cracking in PWR steam generator tubes, girth welds, an_d feedwater nozzles..The need for revising the ASME fatigue design curves to include environmental effects should be evaluated.

High-cycle thermal fatigue has caused crack initiation at the inside radius of the BWR feedwater nozzles and in the PWR high-pressure injection lines.

The field-fatigue failures have revealed several sites suscci,tible-to fatigue damage _ that were not originally considered vulnerable to fatigue.

Examples of

)

such sites include welds and elbow base metal in.PWR surge lines, safety injection lines, and residual heat removal 7 piping, and. include PWR. steam generator girth welds.

Base metalt sites -in elbows are susceptible to. fatigue damage because 'of the ovalization of the elbow cross section caused by inplane 1

- bending; the most susceptible site is' the inside surface.of the flank of an

_el bow. Fatigue test results for pipe bends and finite element analyses for surge.

~

3 lines subject to thermali stratification have identified elbow base metal as a susceptible site for. fatigue damage.

Current inservice _ inspection requirements concentrate on the inspection of welds; however, asLdiscussed_--above, several-field failures have occurred in-the base metal-(away from welds) not_ included in the inservice inspection program. ASME Coden Section XI requirements - might be expanded to - include sites where high I Work sponsored by the U.S. Nuclear Regulatory Commission. Office of Nuclear Regulatory - Research, under 00E Contract No. - DE-AC07-761001570; Dr. G. H. Weidenhamer, Technical Monitor.

l 21 a

_.- ~

i fatigue usage is predicted (including base metal sites) or where there has been a history of failures._ Such sites would include surge line elbows and feedwater

)

nozzles, -Inservice inspection requirements in-some other countries include examination of. susceptible base metal sites.

For example, the inservice inspection requirements ~ for the newest generation of nuclear' power plants in Germany include inspection of the elbow base metal, inservice inspection experience indicates that the detection of thermal fatigue 1

cracks is difficult because these cracks generally are tightly closed during the j

inspection period when pressure and thermal stresses are reduced.

The conventional inservice inspection techniques and procedures satisfying the ASME Code minimum requirements are inadequate to detect the presence of such tight

cracks, inspections with conventio,al ultrasonic methods were unable to detect 1

a through-wall crack on the Farley safety injection line.

Reliable inservice.

inspection for the fatigue cracks may be performed during hydrotests when the cracks are opened sufficiently to be detected, Advanced methods, such as the time-of-flight diffraction technique, need to be developed and field tested to inspect sites susceptible to fatigue cracking. Acoustic emission monitoring for fatigue crack growth needs to be tested in the field.

Uncertainty in the amplitude of the stressors and limited access for inspection at certain susceptible sites have resulted in a need for the developmeat of on-

-line fatigue _ monitoring techniques. - Utilities are applying these techniques to more accurately determine the severity and numbers of cycles for the transients that contribute to fatigue damage at susceptible locations. Further development of these techniques will be of potential benefit in better estimating the-fatigue usage at critical sites, in supplying useful information in making' inspection and repair / replacement decisions, and in modifying operating procedures to mitigate fatigue damage.

The present generation of nuclear plants was not instrumented to acquire all the. information needed for fatigue monitoring; thus, indirect measurements from_the existing plant instrumentation must be used. -Some newly constructed nuclear power plants in Germany, France, and Japan are being instrumented with additional local sensors for the express purpose of fatigue monitoring.

This should be considered in the design and construction of the advanced LWRs, Understanding the stressors that cause fatigue damage can-lead to the development of mitigaticr> techniques, which include modified designs and changes in operating procedures.

For example, the cladding.has been removed from the inside radius of BWR feedwater nozzles to. eliminate stresses caused by differential thermal expan-ion between the cla6 Jing and the base metal.

Some plants now have preheating tanks to raise the temperature:of auxiliary feedwater closer to that of steam generator coolant, thus mitigating the thermal shock loads. Procedural requirements also can be : implemented, such as limiting the differences in pressurizer and reactor coolant system temperatures to no more than 200 F during heatups -and cooldowns to reduce the effects of thermal. stratification in surge

lines, 22 L

. ~

f A COMPRulDlSlyl APPROACH TO MANAGE AGl!LG.

JJN NUCLEAR SULYLCLWATER SYSlutS I

A. B. Johnson, Jr. and D. B. Jarrell Pacific Northwest Laboratory J. J. Burns U.S. Nuclear Regulatory Commission Pacific Northwest Laboratory,I*) in support of the Nuclear Plant Aging Research (NPAR) Program, has conducted aging assessments of service water systems (SWSs) in nuclear plants.

The objectives of the assessments are to provide guidance to manage and mitigate aging of SWSs and to address system-specific options for water chemistry control, cleaning, and surveillance.

The scope of the study encompasses the following elements:

review of databases for SWS failure histories interfaces with experts in SWS chemistries and corrosion visits by PNL teams to three nuclear power plants for comprehensive assessments of SWS age-related degradation participation in NRC case studies of SWSs at two nuclear power stations reviews of relevant literature publications that summarize results of the NPAR FNS aging assessment.

SWS designs at nuclear plants operating in the United States include three general types:

open type retirculating type closed type.

These designs differ in functions and capabilities for water chemistry control, cleaning methods, and, in some details, surveillance.

The following characteristics regarding SWSs should be considered when developing a compre-hensive approach to manage the effects of age-related degradation:

Major portions of the systems are critical to safety, including service to key safety-related equipment.

The systems are relatively large and have numerous components (e.g., ~40 or more heat exchangers, four to ten major pumps, and over 200 valves).

(a)

Pacific Northwest Laboratory is operated for the U.S. Department of Energy under Contract DE-AC06 76RLO 1830.

Work is conducted under NRC FIN B2911.

23

1 l

i The systems are subject to a variety of aging mechanisms that impose challenges to determine causes of degradation and to select appropriate counter measures.

A given system is subject to a variety of service conditions; for example, some surfaces are wetted, some are exposed only to air, and some are expc'ed to soils; a given component may operate stagnant, under flow, or intermittently; wetted surf aces may be subject to abrasive particulates, biota, and detritus.

l Many of the corrosion-prone surface areas are not readily accessible for visual inspection.

I The vulnerabilities to degradation are system-specific and component-s;:scific, depending on the materials, designs, water compositions, etc.,

of each system.

The degradation rates and mechanisms have seasonal variations.

Unexpected phenomena, such as encroachment of previously undetected biological species, can occur.

Effective SWS management should begin with the premise that active strategies involving timely monitoring and control are preferable to reactive measures associated with recovery from highly degraded states.

The key elements of effective SWS management are as follows:

A.

sound design B.

appropriate materials selection C.

system-specific water chemistry control D.

systematic monitoring, inspection, and surveillance E.

system-specific cleaning on a planned schedule F.

systematic and timely maintenance (Note: 0 and E could be regarded as elements of maintenance.)

G.

Attention to operating history, including trending.

Decisions concerning corrective actions should a based on analysis of the root cause(s) of observed degradation, proposed resolutions, and the effects of the proposed resolutions on the total system. The need for corrective actions may arise from errors in design or materials selection, from changes in operating conditions after the design was completed, or from degradation anticipated during ooeration.

24

ASSESSMENT OF DIAONOSTIC METHODS FOR DETERMINING DEGRADATION OF MOTOR-OPERATED VALVES

• H. D.

Haynes Oak Ridge National Laboratory W.

S.

Farmer U.

S. Nuclear Regulatory Commission

SUMMARY

Motor-operated valves (MOVs) are located throughout nuclear power plant fluid systems. Their f ailures have resulted in significant maintenance ef forts and, on occasion, have led to the loss of operational readiness of safety-related systems. Thrt Oak Ridge National Laboratory (ORNL) has carried out a comprehensive aging assessment of MOVs in support of the Nuclear Plant Aging Research (NPAR) program. This paper providos a summary of the ORNL MOV aging assessment with emphasis on the identification, evaluation, and application of MOV monitoring methods and techniques.

Beginning in 1985, the diagnostic information available from many MOV measurable parameters was evaluated by ORNL using MOVs that were mounted on test stands.

Those parameters included valve stem position, valve stem velocity, valve stem strain, torque and limit switch actuation times, internal and external motor temperatures, vibration, torque switch angular position, and motor current. Those tests led to the conclusion that the single most informative MOV measurable parameter was also the one which was most easily acquired, namely the motor current. Motor current signature analysis,MCSA) was found to provide detailed information related to the condition of the motor, motor operator, and valve across a wide range of levels from mean values and gross variations during a valve operation to information which characterizes transiente and periodic occurrences.

A detailed discussion of the application of MCSA.to MOVs is provided, including examples of time waveform analysis, f requency spectral analysis, and additional techniques found useful in determining MOV performance and condition.

h ha part of the MOV aging assessment, several tests were carried out by ORNL on MOVs having implanted defects and degradations. Tests were also carried out on many MOVs located within a nuclear power p'. ant. In addition, ORNL participated in the Gate Valve Flow Interruption Blowdown Test program carried out at H ic f

Laboratories in Huntsville, Alabama. Recults from all of these teste are summarized in this paper and several celected examples are given.

Other areas covered in the paper include descriptions of relevant regulatory

. issues and activities, other related diagnostico research at ORNL (including the application of MCSA to other equipment), and interactions ORNL has had with outside organizations for the purpose of disseminating research results.

=

Research sponsored by the office of Nuclear Regulatory Research, U.S.

Nuclear Regulatory Commission under Interagency Agreement DOE 1886-8082-8B with the U.S. Department of Energy under contract No. DE-AC05-840R21400 with Martin Marietta Energy Systems Inc.

me

.4,,w.s e.

o,i m

pen ennhores by e arisecer of te U S.

Guvernmere wt(ar ear. West h M-ACIE44CFJ M0G W #8 ua 25 G-~'**

'*=a' i

enveny4we toesus to pm'en er evevoe"ma

,l Ww pHoheheel Opem of Wee -

1 er l

I asaw oehere to de eso, far SE Gouesmere pwpuses

• LIFE Tt' STING OF A LOW VOLTAGE AIR CIRCUIT BREAKER

• Mano Subudhi Brookhaven National Laboratory Upton, New Yerk 11973 and Satish Aggarwal 5

U.S. Nuclear Regulatory Commission Washington, l' 20555

SUMMARY

A DS-416 low voltage air circuit breaker manufactured by Westinghouse was mechanically cycled to identify age related degradation in the various breaker subcomponents, specifically the power-operated mechanism. This accelerated aging j

test was performed on one breaker unit for over 36,000 cycles.

Three separate pole shaf ts, one with a 60-degree veld, one with a 120 degree weld, and one with a 180-degree.reld in the third pole lever were used to characterize cracking in the welds. In addition, during the testing three different operating mechanisms 9

and several other parts were replaced as they degraded to an inoperable condition.

Pole shaf ts used in this test program were found to have substandard welds.

Examination of a fractured trip shaft lever suggested inadequacies in electroplating techniques. Newly purchased reset springs had r. harper bends at the neck of the hooks than an older design, which led to early spring failures.

Testing of the hardness of the oscillator surface showed a 30% reduction for the newly ptocured units.

Wear, fracture, distortion, and normal fatigue dominated the aging process with wear being the largest contributor.

Excessive wear was evident in the ratchet wheel, holding pawis, oscillator, drive plate, motor crank and handle, cam segments, main roller, and the stop roller.

Structural components and contact assembly parts indicated that there was little aging due to mechanical cycling. A pole shaft with a reduced size weld could fail at a cycle as low as 3000.

The ultimate life of various breaker parts was found to be 10,000 cycles, except for the newly procured reset - spring, which was 2000 cycles.

One commercial grade lubricant was found to perform better than those recommended.

The current plant. maintenance practices need to incorporate the experience of aging problems associated with the power-operated mechanisms. The scheduling of the breaksr maintenance should be dependent primartly on the number of cycles experienced by the breaker, with some consideration given to time in service.

Tha maintenance and manufacturing recommendations, obtained from the test results, should help mitigate aging problems. When procuring a new breaker or spare parts, careful attention should be given to their design adequacy. For the

' indicator and reset springs, it is recommende d that a smooth transition bend should be made instead of a sharp bend to minimize the damage to the surface, and

• Work performed under the auspices of the U.S. Nuclear Regulatory Commission.

26

to reduce the ensile stresses on the insido surf.

of the bend area of the ont books.

For the pole pin connectivir, Ihase A of t_.

breaker contact s, sinoot h corners should be machitied, free froin surface def ects, linproper velding, practices at the pole shaf t levers should be avoided.

Assuming; a factor of safety of 2, the life of a DS 416 (or "t 206) breaker is estimated to be 5000 f

cycles.

Based on an assuniption that a breaker, such as reactor trip, is typit. ally subjected to 250 cycles annually; this translates to a breaker life of

(

20 years, I

f 4

(

l=

i l

i 1

5 f

h j

(

i b

+

i 27

-i

,-__.,_.4.,,,,;.,

-.,.,_.,..._...,...,-J.,,,-,-,,

,-.....,,_J..-_...,,.,__,..._..,..._......_._-_.

Aging and Low. Flow I)egradallon of Ausiliary Feedwater Pumps' M. L Adams Case Western Reserve University i

for Oak Ridge National Laboratory 4

NRC Technical Monitor: Dill Fanner

SUMMARY

Auxiliary feedwater (AFW) pumps are used in safety.related AFW Systems at pressurized water reactor plants. The function of AFW pumps is to deliver water from either a l

condensate storage tank or, as a backup. from the emergency service water system, to the i

stearn generators. The water that is pumped to the steam generators is evaporated, thereby i

removing decay heat from the reactor coolant system.

1

)

^

The pumps are automatically staned in response to several emergency conditions, such as low steam generator level, a safety injection signal, and emergency bus undervoltage.

llowever, many plants also use the pumps in support of nonnal shutdown and startup requences, since the main feedwater system pump capacity greatly exceeds demand during these conditions. The other prinelpal service seen by the pumps is during testing.

The AFW pumps are multistye (normally 5 to 9 stages) high. head centrifug.J pumps, I

normally driven by motors or 'mbines. Rotating speeds for the motor-driven pumps are l

nominally 3550 rpm, while those for turbine-driven pumps are routinely closer to 4(KX) i rpm. Rated pump deliveries range from roughly 200 to 1200 ppm.

Much of the operation of AFW pumps is at low How conditions. Once the reactor has been shut down, the reactor decay heat generation rate and the attendant AFW pump flow a

requirement drop exponentially. For example, if full pump capacity is required to remove all decay heat at ten minutes after a reactor shutdown,just over a third of the pump capacity would be required for heat removal after five hours. Of course the heat removal requirements during reactor stanup following an outage are insignincant relative to typical pump capacity. As a result, much of the pump operation in support of these stanup and shutdown evolutions is essentially at minimum Dow.

' Research monsored by the Offiee of Nuclear Regulatory Research, U. S. Nuclear Regulatory Commission under Interagency Agreement DOE 1886-8082 8H with the U. S. Department of Energy under contract No.

DE AC05 840R21400 with the Martin Marietta Energy Systems,Inc.

28 e

-vp -e.

-,.,r,-v w w.

-,,r w---,m

.--2,e y--r.:,=r.----w.

y w,-e.w-wwswww-?+rw--re--rw=----,-w-wv.re'e-,w,-rc-v,-e,--e~re--mww-se e

e-=.-wwwe w

w--w w t=aw

-+=+--,e

l Many of the ARY pumps are provided with relatively low minimum now line capacities, typically in the range of 5 to 10% of best ef ficiency point now. The pumps are normally tested under minimum flow conditions. Testing is conducted, per plant Technical Specifications, on a monthly or quarterly basis. In addition, automatic start testing is nonnally required to be conducted every eighteen months The minimum now provisions for many of these pumps me not adequate to support continuous operation. The installed mininow lines were, in most cases, intended by the p"mp manufacturers to te used just during pump starting and stopping. Ilowever, the required testing as well as the necesshy of using these pumps for plant startup and shutdown support has resulted in a considerable amount of operation at low flow conditions. Operation at low now results in accelerated wear of the pumps due to vibration associated with the hydraulically unstable conditions. The wear can be manifested in a

~

number of ways, such as impeller or diffuser breakr,c, thrust bearing and/or balance device failure due to excessive loading, cavitation oamage on suction stage impellers, increased seal leakage or failure, seal injection piping failure, shaft or coupling breakage, and rotating element scirure.

This paper discusses pump design, historical operating experience, and testing and inspection methods. Examples of operating problems that have tren experienced and potential design malineations are provided.

29 l

Aulng Evaluation of Nuclear Plant RTDs and Pressure Transmitters H. M. Hashomlan Analysis and Measutomont Servicos Corporation AMS 9111 Cross Park Drivo vnoxvillo, TN 37923

\\

(015) 091 1766 AJJTRACT Rosistance Temperatuto detectors (RTDs) and pressuto, level, and flow transmitters provido almost all the v#al signals that are used for the control and safoty of nuclear power plants. Thoroforo, it is crucial to ensure that tho performanco of thoso sonsor romain adequato as they ago in the process under normal operating conditions.

Two comprohonsivo roscarch projects woro conducted for the NRC to evaluato tho offects of normal aging on calibration stability and rosponso timo of RTDs and pressure transmittors of the typos used for safoty rolated measutomonts in nuclear power piants. Each project was l

conducted over a throo year period. The projects involved laboratory testing of representativo l

RTDs and pressure transmitters aged in simulated roactor conditions. The main purpose of thoso l

projects was to establish the degradation rato of the sonsors and use the information to l

dolormino if the current testing intervals practiced by the nucloar power industry is adoquate for l

management of aging of the sonsors. The results havo indicated that the curront nucloar industry practico of testing the response timo and calibration of the sensors onco ovory fuel cycle is adequato. This is provided that all the sofoty-related sensors are testod for both calibration and response timo as opposed to testing one out of ovory two or four of the redundant sensors.

in addition to working to identify degradation ratos and testing intervals of the sonsors, a few outstanding issues rogarding the performanco of nuclear plant RTDs and pressure transmittors woro addressed. Those included a comprehensivo assessment of the acceptability of the cross calibration method for on-line testirg of calibration of installod RTDs, the oilloss phenomenon in Rosomount pressure transmitters, and the offects of sensing lino blockages on the overall responso timo of nuclear plant pressure sonsing systems, For both RTDs and pressure transmitters, the LER and NPRDS databases wore searched and analyzed to au0mont the experimental data wo generated by laboratory tosis.

30 t

POWER REACTOR EMBRITTLEMENT DATA BASE (PR-EDB):

USES IN EVALUATING RADIATION EMBRITTLEMENT OF REACTOR VESSELS

• F. B. R. Kam, F. V. S t al linann, and J. A. Vang Oak Ridge National Laboratory Investigations of regulatory issues such as ve.asel integrity over plant life, vess.el failure, and sufficiency of current Codes, Standard Review Plans (SRPs), and Guides for license renewal can be greatly expedited by the use of a well designed, computerized data base Also, such a data base is essential for the validation of embrit tleinent prediction inodels by researchers.

The PR-EDB is such a comprehensive collection of data for U.S. power reactors..

The cottpilation of the Test Reactor Embrittlement Data Base (TR EDB) is in progress and will supplernent the data in the PR EDB.

More analyses of test and power reactor results are needed to estahl'sh the applicability of tout reactor data to power reactors.

" User-friendly" software programs to access, process, and manipulate the data have been written and further developments are expected.

A list of studies using the PR EDB f ollows:

1.

Analysis of the correlation monitor inatorials (A302B and A5338) in surveillance capsules of commercial power reactars (NUREC/CR 4947);

2.

Differences between the measured transition temperature shif ts (ARTnor) and the v tues predicted by Regulatory Guide 1.99 (Rev. 2) for base and weld materials; 3.

Comparison of ARTso7 and the reduction in the Charpy V upper r.helf energy (AUSE) for specimens fabricated in the longitudinal orientation (LT) with those in the transverse orientation (TL);

4.

Comparison of USE from LT specimens reduced to 65% of their value with the USE values from TL specimens; 5.

Comparison to see if Regulatory Guide 1.99 (Rev. 2) underpredicts the Charpy shift for low copper material when the phosphorus content is hirlt; 6.

Comparison of the two standard deviation value (20) for the residual

[ observed value of ARTuny minus value calculated by Regulatory Guide 1.99 (Rev. 2)} from Cuthrie's data base and the PR EDB;

• Resear*h sperisured by the Of fite of Nuclear Regulatory Researth, Divitic.n of Ingineertng, U.S.

Nuclear Regulatory Cc m assion under interagency Agreement DOE 16 M80415B with the U15. Department of rnetsy under contract De-AC05-640R21600 with Martin Mariette f.norgy rystems, Inc.

The suleitted rLanuscript has been authoreJ by a contractor of the U,$. Goverteent under t.ontract No.

DE-Ac05-040K21400. Accordingly, the U.S. Goverreer.t retains a nor. exclusive. royalty +f ree 115 ense to publish cr reg roduce the }vblished form of this contilbutton, or allow others to do so, for U.S. Goverreent gurgears.

31

7.

Coniparison of the Regulatory Guide 1.99 (Rev. 2) toodel with the French F1H inodel in predicting ARTwer for base materials and welds; 8.

Con >parison of the TANil fit ting program in PR l'DB with one from the Elcetric Power Research Inst itute (EPRI);

9.

Analysis of instrumented Charpy data based on model by D.

Pachur;

10. Study of plate teaterials siin11ar t o the Yankee Rowe Reac t or vessel and t o staterials in their test reactor experinant al pro 6 tam; and

11. Generation of a data file that has similar radiation and annealing environtnent s as the Yankee Rowe Reactor to test model prediction.

32

Aging impact on the Safety and Operability of Nuclear Reactor Pressure Yessels' W. ii. Fennell Oak Ridge National Lateratory Oak Ridge,TN 37831 Structural integrity of the reactor pressure vessel must be assured throughout the operating life of a nuclear power plant in order to maintain the capability to cool the nuclear core. This assurance is achieved by requirin; that the reactor vessel maintain specified fracture-prevention margins throughout its operating ife. Fracture prevention margins are calculated using fr. - *e mec : nudes technology in conjunction with a material property known as fracture toughness.

Fracture tougfmess is a measure of the ability of a material containing sharp edged cracks to sustain stress. Fracture-mechanics based structural integrity assessments differ therefore from those obtained from the more familiar stresostrength analysis in that they account in a quantitative manner for the effect of cracks, which are unavoidably present in all engineering structures.

l Irradiation damage is the aging mechanism of dominant concern in pressurized-water-reactor pressure vessels. Irradiation exposure causes atom displacements in the vessel material lattice structure which have the effect of increasing its strength but decreasing its ductility and fracture toughness. It follows therefore that fracture prevention margins will progressively decrease as the vessel material absorbs increasing amounts of irradiation damage throughout its I

operating life.

Regulatory requirements limit the permissible accumulation of irradiation damage in the material of a given reactor vessel. Irradiation damage limits are set such that required fracture-prevention margins are maintained throughout the nuclear plant licensed operating period. The regulatory requirements are based on fracture mechanics technology and utilize matenals aging data drawn from mandatory reactor vessel irradiation damage surveillance programs. They address both normal operation of the reactor system and potential accident loading. For nonnal operation, the regulatory requirements dictate that plant technical specifications be periodically adjusted to

,reclude operating conditions which could reduce the fracture prevention margins. For accident oading, they set regulatory limits on the acceptable level of irradiation damage in the vessel material. They also define the scope and acceptance criteria for fracture margin assessments which must be performed to suppon any proposal for continued operation of the plant once the regulatory irradiation damage limits have been exceeded, in recent years it has become evident that a number of nuclear plants will exceed the regulatory limits on irraciation damage to the reactor vessel material before the end of th(it current licensing period. One result of this development is that a number of nuclear industry organizations have gained experience in the application of fracture margin assessment technology. This experience has resulted in the identihcation of a number ofissues with the technology in its present form. Data from irradiation testing programs, operating plant surveillance programs and large scale fracture technology validation tests have identified additional issues. The NRC funded Heavy Section Steel Technolo gy program at Oak Ridge National Laboratory is performing the researc' required to resolve t aese issues and further develop and refine the fracture margin assessment technology.

"Research sponcored by the Orfice or Nucicar Regulatory Research.

U.S. Nuclear Regulatory ConrJssion urder Interagercy Agreewnt 1886-8011-93 with me U.S. Department of Energy urder Contract DE-ACOS-840R21400 with Martin Marietta Erergy Sysiums. Inc. The submitted freuscript has been authored by a contractor cf the U. S. Coverrnnt under Contract DE AC05 E.OR21400. Accordingly, the U. S. Covertrent retains a nonexclusive, royalty free license to publish or reproduce the published torm of this contribution, or allow others te do so, for U. S. Coverment purposes.

33

This paper will present a brief overview of the current status of fracture prevention regulatory requirements and he associated fracture-margin assessment technology. Issues identified with the technolo.!y will be reviewed. Itesearch pmgrams implemented to resolve the issues will be described.

!otential impacts of this ongomg research on the fracture margin assessment process will be discussed.

4 4

i 1

1 l

34

e The Application of Probabilistic Fracture Analpis to ltesidual 1.lfe I: valuation of Embrittled lleactor Yessels' T.L Dickson Oak itidge National laduratory Oak itidge,TN 37831 F. A. Simonen Pacine Nonhw est I.aboratories Richland, WA 99352 Irradiation damage is the aping mechanism of dominant concern for pressurized. water reactor (PWR) pressure vessels. Lumulative irradiation exposure has the effect of making the vessel material more brittle (decicased fracture toughness) and therefore more susceptible to cleavage fracture. For such embrittled vessels, pressurized-thennal shock (IrfS) is the major challenge confronting vessel inteprity. Concern with PTS results from the combined effects of (1) pressure and thermal shock loadmgs,(2) embrittlement of the vessel material due to irradiation, and (3) the possible existence of sham, crack-like defects at or near the inner surface of the vessel.

The PTS issue has been under investigation for many years, hiost of the early PTS analyses were of a conservative detenninistic natme. In an effon to establish more realistic limiting values of vessel embrittlement, the NRC funded the Integrated Pressurized Thennal Shock (IPTS)

Program, which developed a comprehensive probabilistic approach to the PTS vesselintegrity issue. A major element of the probabilistic approach is perfornung probabilistic fracture mechames (PFhi) analyses to predict the probability of vessel failure assuming a specined PTS event occurs at a speciGed time in the operating life of the plant. This paper describes the application of PFhi to predict the residual life of embrittled vessels in accordanse with regulatory requirements.

Current regulatory requirements for vessel integrity are based on the probabilistic methodology and Regulatory Guide 1.154 which provides guidance to utilities on how to perfonn PTS probabilistic vessel integrity evaluations. Regulatory Guide 1.154 teferences OCA P and VISA Il as acceptable PFht computer codes for performing the probabilistic fracture mechanics portion of the evaluations. These codes predict the increase in vessel failure probability that occur as the vessel material accumulates irradiation damage as a function of time. Such results, when compared with limits of acceptable failure probabilities, provide an estimation of the residual life of a specific vessel. These codes can be used to evaluate the benefits of plant-speciGe mitigating actions which have the potential to extend the residual life of a vessel.

• Research aponnoted by tre Of fice of Nuclear Regalatory Pescarch.

U.S, Nuclear keguiet ory Omission u"4cr interagency Agreement 18b6 !s0ll-93 with the U.S. Departrent o r F.ne r gy unde r contract DS ACM MOk2:400 with Martin Marietta frorgy Systwa. inc.

The trAcitted muscript has been authored by a contractor of the V, S. Govertrent t;nder Contract DE ACC& MOR?l400. Accordirigly. the U. S. Goverrront reteins a nonexclusive, royahy f rec licenne to ;>ublich or teproduce Om published form of t his contribmion, or allow others to do s.a. for U. S. covem~cnt pur;+ows.

35

Managing Irradiation Embrittlement in Aging iteactor Pressure Vessels' W. R. Corwin Oak Ridge National Laboratory Oak Ridge, Tennessee 37831 Maintaining integrity of the reactor pressure vessel (Rl'V) in light-water cooled nuclear power plants is crucial in preventing severe accidents and their potential for major contamination releases.

This requires a quantitative understanding of irradiation induced degradation of the RPV's fracture resistance, a methodology for assessing the impact of reduced toughness on integrity, and a means of deciding under what conditions the vessel can continue to operate safely. For reactor vessels recently fabricated from radiation-damage resistant steels or with very thnited service, there is little concern over challenges to their integrity since without significant radiation damage, it is virtually impossible to postulate a realistic scenario resulting in RPV fracture. Ilowever, for aging reactors containing pressure vessels fabricated before the mid-1970s, when metallurgical controls were instituted to control embrittlement in vessels, the irradiation induced loss of fracture resistance could unacceptably compromise vessel integrity under severe loading conditions.

Unfortunately, with the United States having pioneered civilian nuclear power, many of our older reactors fall into this later category. It is therefore imperative to understend and predict the effects of irradiation on RPV steels to ensure the continued safe operation of these aging reactors. For this reason, the licavy Section Steel Irradiation (llSSI) Program has been established by the U.S. Nuclear Regulatory Commission (USNRC) to provide a thorough, quantitative assessment of the effects of neutron irradiation on the material behavior and, in particular, the fracture toughness properties of typical pressure vessel steels as they relate to light-water reactor pressure-vessel integrity. Results from the llSSI studies provide information needed to aid in resolving major regulatory issues facing the USNRC, which involve RPV irradiation embrittlement, such as pressurized-thermal shock, operating pressure-temperature limits, low temperature overpressurization, and the specialized problems associated with low upper-shelf (LUS) welds. The results of these studies, when used in conjunction with results from other USNRC-sponsored programs that provide comparable info mation on the detection and evaluation of flaws and overall

'Research sporsored by the Office of Nudear Regulatory Research, U.S. Nuclear Regulatory Commission, under interagency Agreement DOE 1886410941. with the U.S. Department of Energy under ct otract DE AC05-840R21400 with Martin Marietta Energy Systems, Inc.

The submitted manuscript has twen authored by a contractor of the U.S. Government under contract DE-AC05440R21400. Accordingly, the U.S. Covernment retains a nonexclusive, royalty-frce license to pubilsh or reproduce the published form of this contnbution, or allow others to do so, for U.S. Government purposes.

36

wi mm I

fracture assessment methodologies, provide guidance and bases for both managing irradiation induced embrittlement in aging vessels and evaluating the potential for their extended life operation.

hiajor program elements within the llSSI Program include:

experimental investigation and verification of the irradiation induced loss of fracture resistance in critical reactor vessel materials, with an emphasis on applicability to the thick sections used in Rl'V construction; experimental verification and expansion of existing data on annealing recovery and subsequent reembrittlement in critical pressure vessel materials to provide a validated basis for remedial actions for severely embrittled vessels; coordinated, advanced microstructural examinations and physically based theoretical model developmer.t of controlling microstructural mechanisms to provide improved predictions of macroscopic embrittlement; verification of predictions of radiationynduced damage by examination of materials exposed during actual service; and maintaining the supply of correlation monitor material used for validating results of world-wide irradiation surveillance programs. The principal materials examined within the 115S1 l'rogram are high-copper welds since their postirradiation properties are most frequently limiting in the continued safe operation of commercial RPVs. The irradiation induced shif ts and changes in shape in fracture toughness were examined for two high copper welds. The irradiation-shifted fracture toughness fell slightly below the AShiE K]c curve even when it was shifted according to Revision 2 of Regulatory Guide 1.99, including its margins, indicating that the current method for assessing fracture toughness reductions may be nonconservative. The high copper beltline weld removed from the hiidland reactor is being examined to establish the effects of irradiatien on a commercial 1.US weld. A wide variation in the unirradiated fracture properties of the hildland weld, with values of RTNDT ranging from -a to 20 C and copper contents from 0.16 to 0.16 wt

%, were determined. Experiments have been initiated to examine the irradiation embrittlement of the LUS weld as well as provide initial data in the annealed and reirradiated conditions Annealing experiments on critical high copper materials

~

have been initiated to provide the first quantitative comparison of Charpy V notch versus fracture toaghness recovery and reembrittlement. Embrittlement modelling studies have shown that the dose required for point-defect concentrations, which contribute to irradiation embrittlement, to reach steady state values can be comparable to component or irradiation experiment lifetime. Thus, embrittlement models that sely on the assumption of steady state at these conditions are not valid.

Consequently, simple direct comparisons of embrittlement generated under different rates of exposur 7 (e.g., test reactors verses power reactors) may be misleading because the relative state within the initial defect production transient will likely be different.

Since the effect of displacement rate is different in the transient and steady-state regimes, unqualified data extrapolation from test reactor irradiations may cross mechanism boundaries and lead to poor estimates of material response at low displacement rates.

37

DETECTING AND MITIGATING AGING IN COMPONENT COOLING VATER SYSTEMS

• Robert J. Lof aro Brookhaven National Laboratory Upton, New York 11973 and Satish Aggarwal U.S. Nuclear Regulatory Conunission Washington DC 20555

SUMMARY

The cortponent cooling water (CCW) systern is one of many systems that is litpo r t ant for safe operation of nuclear power plants.

In a research program sponsored by the U.S. Nuclear Regulatory Conunission (NRC), the CCW system has been studied to determine how aging "fects its perforitance and reliability. The study was pe r fortrie d in two phases and included extensive analyses of data obtained from national data bases, as well as data obtained frotn actual plant visits. This paper discusses how the results of those analyses can be used to help detect and mitigate the affects of aging in CCW systems.

The function of the CCW system is to rettove heat from various loads throughout the plant and discard it to an open loop cooling system, such as the service water system. The loads serviced by the CCW system can be safety related or non safety related, such as the reactor coolant pump reals, the shutdown heat exchangers, the residual heat removal heat exchangers, and the safety injection purop s.

Due to the diversity of the loads dependent on it, the CCW system is continuously operating, and is required during normal, as well as off normal plant operation.

Therefore, the affects of aging must be properly inanaged to ensure safe plant operation in later years.

l In phase 1 of the CCW system aging study an analysis of past operating experience showed that the CCW components are susceptible to aging degradation, and that this degradation leads to an increase in failure rate as the components age. Of the failures reviewed, over 70% were related to aging degradation, with the dominant cause of failure being " normal service."

The dominant failure mechanism was " wear," which is consistent with the high percentage of failures attributed to aging. The components having the largest number of failures were valves, followed by pumps, instrumentation, and heat exchangers.

1 Using t irne -dependent failure rates calculated from the

data, a

probabilistic 'isk assessment (PRA) analysis was done for a typical CCW system design. The re,ults showed that if component failure rates increase with age the unavailability of the system will also inctease.

Since the CCW systam is important to safety, this could lead to an increase in plant risk.

These i

findings clearly show that proper detection and mitigation of aging degradation should be an important part of daily plant operation.

t

1. Work performed under the auspices of the U.S. Nucicar Regulatory Commission, 38

To detericine the snost of feet ive methods of managing aging, a 5.econd phase 2

of the CCW system aging study was pe r f o rtne d in this part of the study inspection, surveillance, monitoring, and snaint enance (I SMM4) practices were investigated.

Information on ISMMi practices currently used at plants was obtained f rom a survey, along with actual plant visits and personnel interviews.

In addition, various advanced practices were ident if ied through literature searches and discussions with component manuf acturers. The findings provided an excellent overview of what methods are available to properly control aging t

degradation.

The information on currently used ISM &M practices showed that there are two ca egories; basic practices, which are typically required by codes or plant technical specifications, and supplemental practices, which are selected based on particular plant operating characteristics and environroent.

An effective ISMMi program requires a combination of basic and supplemental practices to ensure that at least one method is in place to detect and mitigate each of the common aging, mechanisstn that may lead to component failure As an aid in evaluating a plant.'s ISMMi programs, the various practices identifled in this

.itudy were correlated with the aging mechanisms they can detect and/or mitigate, and the results were tabulated for each of the maj or corrponent s.

These tables are included in the full paper.

REFERENCES

1. liiggins, J., et al., " Operating Experience and Aging Assessment of Component Cooling Water Syst ems in prest.urized Vater Reactors," NURI:G/CR 5052, July 1988,

2. 1ofaro, R.,

et al.," Aging Assessment of Component Cooling Water Systems in pressurized Water Reactora phase 11," NUREG/CR 5693, To Be published 1991.

i I,

39

ifMBER AMNG AS.Sf3Mit(I I

D. P. Brown Lake Engineering Company D. E. Blahnik Pacific floi hwest Laboratory J. J. Burns U.S. tiucicar Regulatory Commission The Pacific florthwest Laboratory (Pl4L),M in support of the fluclear lant Aging Research (tiPAR) Program, has conducted aging assessments of snubbers used in commercial nuclear power plants.

The objectives of the assessments were to characterize snubber aging and to provide a basis ior managing and mitigating its effects.

Phase I of the study was completed and the results are published in f40 REG /CR-4279.

Phase 11 was also completed by Pill and its subcontractors, Lake Engineering Company (LEC) and Wyle Laboratories.

During Phase II, we conducted a feasibility study (NUREG/CR-5386) and identified specific aging research needs.

The remainder of Phase 11 involved in plant research with several nuclear plant utilities, where we interviewed plant maintenance and engineering staff, and analyzed plant oper-t ating data.

Thirteen plants at eight different sites were visited during a three-month period.

Snubbers at five of the sites were primarily mechanical; snubbers at the remaining three sites were primarily hydraulic.

Laboratory research was also conducted at the LEC facility, and general information per-taining to aging was obtained from LEC's files, for this sit.dy, we defined snubber aging as " showing the effects of time and use on the physical charac-teristics of a snubber." By distinguishing between snubber failures related to aging and failures related to nonaging causes, we concluded that approxi-mately half of all snubber failures were attributable to aging influences that

(

involve the operating environment (temperature, humidity, etc.). dynamic transients, and vibration.

Nonaging-related failures are associated with j

snubber design or manufacturing inadequacies, improper assembly, or damage incurred during plant construction or during snubber installation.

. Heat, vibration, and moisture can degrade the performance of mechanical snubbers by incrcasing drag and breaFaway forces, and by changing activation acceleration thresholds.

Data from mechanical snubbers in one piant indicated a slightly increasing trend in drag force with service time.

For hydraulic snubbers, high temperatures in isolated operating areas can rapidly degrade seals.

At one boiling-water plant, the incidence of seal leakage in hydraulic snubbers was higher in the drywell (higher temperature) than in other areas of the plant.

Radiation probably contributes less significantly to seal degra-(a)

Pacific Northwest Laboratory is operated for the U.S. Department of Energy under Contract DE-AC06 76RLO 1830. Work was conducted under flRC

]

FIN B2911.

40

dation than was originally hypothesized.

Aging degradation mechanisms for clastomeric seals include extrusion, embrittlement, aermanent set, wear, and i

adhesion to mating surfaces.

However, most seal leass are not directly attributable to long term service degradation.

fluid leakage in hydraulic snubbers is commonly attributable to leaking fittings, improper assenbly, or design inadequacies.

Several plants have implemented seal life evaluation studies.

Many plants routinely evaluate the causes of snubber failures, often Identifying specific designs or applications that are prone to degradation; augmented surveillance for such applications is common.

However, not all plants have implemented formal service life monitoring programs, for those that have, service-life monitoring generally involves only ensuring that seals in hydraulic snubbers are replaced at prescribed intervals.

Based on results of the Phase 11 investigation, we developed several service life monitoring recommendations for mechanical and hydraulic snubbers:

Information about the snubber operating environment should be identified and can be obtained from the snubbers themscives by visual examination (in-situ and during disassembly), by evaluation of functional test data, by fluid sampling, and by conducting

• hands-on" checks.

In situ monitoring is useful for verifying specific operating con-ditions; various equipment is available for this purpose.

Root causes of failures should be determined.

Determining the root cause of degradation in snubbers removed from service is also advisable.

Data on degradation caused by n service-related influences should be interp.eted separately from data used to verify service life.

Snubbers subjected to severe environmental influences should be identi-fied and managed on a case by case basis.

Service life for the general snubber population (snubbers that do not require augmented mintenance) should be established by trending rele-vant degradation p.rametus such as seal compression set for hydraulic snubbers, or drag force for mechanical snubbers.

Augmented evaluation methods, such as hand stroking, are useful to identify some forms of snubber degradation, such as degradation caused by dynamic load transients.

Service-life projections based on data from snuobers exposed to actual plant operating environments are preferable to analytical service-life projections.

Scheduled maintenance should be based on realistic expectations per-taining to service-related degradation. (Jnnecessary maintenance can increase the potential for snubber failure and result in unnecessary personnel radiation exposure.

41

Managing the Aging of IlWR Control Rod Drhe Systems

• by Rebecca 11. Greene 27,gy yn g Oak Ridge National Laboratory

- - ~. -. - *

• Oak Ridge, TN 37831 8038

$?O," 3 " ".'".* *,.% f,.'

and

'l;;/g,;*,yl0,LT; ma

• v 5

(-'-"'

William S. Fanner, USNRC-RES

• w~"'

5650 Nicholson L:.ne Rockville, MD 20852

SUMMARY

This Phase i Nuclear Plant Aging Research (NPAR) study examines the aging I

phenomena associated with BWR control rod drive mechanisms (CRDMs) and assesses the merits of various methods of "managinf this aging. Infonnation for this study was acquired from (1) the results of a special CRDM aging questionnaire distributed to each U.S. BWR utility,(2) a first-of its-kind workshop held to discuss CRDM aging and maintenance concerns, (3) an analysis of the Nuclear Plant Reliability Data System (NPRDS) failure cases attributed to the control rod drive (CRD) system, and (4) personal information exchange with nuclear industry CRDM maintenance experts. The repon documenting the Grriings of this research, NUREG 5699, will be published this year.

Utilities evaluate the operability of their CRD systems by perfonning individual CRDM scram time testing (as required per plant technical specifications) and weekly to monthly step insenion and withdrawal tests. When a CRDM fails to meet test timing speciGcations or begins to show synotoms such as double notching (erroneously moves two steps instead of one), frequently accomes uncoupled from the control rod blade, exhibits high operational temperatures, or requires excess drive pressure to move, it is I

usually selected for changeout during a plant refueling or maintenance outage. During an outage, utilities typically replace nearly 16% (on the average) of a unit's CRDMs with new or rebuilt units.

Nearly 23% of the NPRDS CRD system component failure reports were attributed to the CRDM. The CRDM components most often requiring replacement due to aging are the Graphitar seals. The predorninant causes of aging for these seals are mechanical wear and thermal embrittlement. Premature aging of these seals is also caused by excessive amounts of crud (din particles, debris, and foreign materials found in the reactor coolant).

This material becomes entrapped between the seals in the CRDM and creates uneven force distributions et the seal's contact surfaces during scrams, causing them to break. Some utilities are vacuuming their reactor vessels in and around the guide tubes during refueling outages to remove and reduce the amounts of crud.

More than 59% of the NPRDS CRJ system failure repons were attributed to components that comprise the hydraulic control unit 0ICU). Each CRDM has a companir n llCU that contains numerous valves which regulate the flow of coolant that controls the movcment of the respntive CRDM. The predominant ilCU valve components experiencing the effects of service wear and aging are the packing, seals, discs, seatr, stems, and c.iaphragms. The llCU valves reponing the most maintenance activity (due to aging) in the NPRDS are the accumulator rutrogen charging cartridge valve, the scram discharge riser isolation valve, the inlet and outlet scram valves, and the scram pilot valve assemblies and their solenoids.

• Research sponsored by the Office of Nuclear Regutatory Research, U. S. Nuclear Regulatory Commission under interagency Agreement DOE 1886 8082 8B with the U. S.

Department of Energy under contract No. DE AC05 840R21400 with the h1artin hlarietta Energy Systems, Inc.

42

The serain water accumulator on pre IlWR-6 IICUs has a carbon steel tank liner w hich has experienced a large arnount of corrosion caused by low Pli water conditions at various plants. Several utility specialists believe that corrosion tir.kes from these af fected accumu' ators have also caused the Teflon seats in the scram valves to erode and plug the companion CRDht's cooling water orifice. To avoid this dep.adation, many utihties are replacing these carbon steel accumulators with a improved component design that features a stainkss steel liner.

Droughout the course of this study,it also became evident that as low as reawnably achievable (ALARA) dose reduction techniques used during CRD system maintenance have become an issue of concern and interest to many utilities. CRDh1 changeout and rebuilding is one of the highest dose, most physically demanding, and complicated maintenance activitics routinely accomplished by llWR utilities. Recent innovations in CRDN1 handling equipment and rebuilding tools have allowed some utilities to make significant reductions in exposures obtained by personnel during the perfonnance of CRDh1 maintenance activities.

• Research sponsored by the Office of Nuclear Regulatory Research. U. S. Nuclear Regulatory Comrnist, ion under Interagency Agreement DOE 1886 8n82 Sil with the U. S.

Departrnent of Energy under contract No. DE AC05 840R21400 with the Martin h1arietta Energy Systems, Inc.

43

llLEQLLOLliO!1110 RING AND TREHQlRG 611L110 TO DIESEL GENERATOR AGING K. R. Hoopingarner Pacific Northwest Laboratory J. J. Burns U.S. Nuclear Regulatory Commission Results of aging studies of emergency diesel generators (EDG) have shown that techniques based V oractical system condition monitoring results could improve average system conditions.

Condition monitoring of about 25 operating parameters of the engine and generator effectively indicates the functional

~ status of important engine components and the location of and when to ap)1y maintenance efforts.

in contrast, intrusive engine inspections and overiauls, basedontimeperiodsalone,(*gendtoreduceengineandsystemreliability.

Pacific Northwest Laboratory under the Nuclear Plant Aging Research (NPAR) program, his completed these aging studies of nuclear service diesel gener-ators.

Based on the results of the aging assessments, it is recommended that a

-reliability centered program for managing emergency diesel generators inte-grating testing, inspection, monitoring, trending and maintenance activities be considered. A reliability centered program will 1) reduce the aging stressors associated with present EDG test requirements, while providing improved confidence in the diesel generator's capability to respond to accident situations, and 2) identify degraded and failing systems and'com-ponents needing replacement or repair before a f ailure actually occurs.

Band on the aging-studies, it is recommended that monthly engine testing be changed to a new approach that involves monitoring and trending. Monthly surveillance testing should involve data acquisition on performance parameters that= indicate trends in component and subsystem cont, tion.

Such data would 3rovide'short-term and long-term information on degraded performance and can

]e used for practical reliability improvements.

Systematic analysis o_f the data offers a basis to improve basic engine reliability and mitigate aging effects.

L(a)_ Pacific-Northwest Laboratory is operated for the U.S. Department of Energy under Contract DE-AC06.76RLO 1830.

Study conducted under NRC FIN

B2911, 44

Aging Studles of Datterlos and Transformers in Class 1E Power Systems' J.L.Edson E. W. Roberts Idaho National En0 ncoring Laboratory i

The Idaho National Engineering Laboratory (INEL) performed aging studies of two components in the Class IE Power System in support of the United States Nucler Regulatory Congnission's (0$NRC) Nuclear Plant Aging Research (NPAR)

Progrw.

Both Phase I and Phase 11 aging research was performed for batteries while transformer research was limited to a Phase I study. The Phase I research consisted of : 1) an identification of the materials of construction, 2) an identification of the aging stressors and the significant aging mechanisms, 3) an examination of operational data to identify the dominant ailure mechanisms that are being observed in nuclear power plants, 4) a review of current maintenance practicos to determine if all significant aging degradation is being detected and managed, and 5) an identification of the needs (if any) for a Phase 11 study.

The Phase 11 study involved performing testing of naturally aged batteries to determine if adequate seismic ruggedness of well maintained, naturally aged, batteries was retained to withstand the most severe earthquakes expected in the U.S.

(he Phase I battery study noted that cracking of the containers and oxidation of the lead are the two most significant aging related mechanisms leading to battery failures.

Oxidation causes the lead to swell, leading to either cracking of containerA or degraded positive plates and decreased electrical capacity.

Oxidation also leads to embrittlement of the lead which, if allowed to continue, will ultimately result in decreased seismic ruggedness.

Rasults of previous battery tests, results of NRC investigations of batteries at nuclear plant with brittle lead components, and the fact that there are no in-plant tests that identify decreased seismic ruggedness pointed to a concern that seismic ruggedness could decrease to less than required to withstand a safe shutdown earthquake (SSE) and yet have acceptable electrical capacity as measured by periodic discharge tests.

Phase 11 seismic tests were conducted on twelve naturally aged batteries obtained from a nuclear power plant..

The batteries were about 13 year old and had been maintained in accordance with the practices specified in IEEE Std 450 and NRC Regulatory Guide 1.1.129, Maintenance, Testing, and Replacement of large Lead Storage Batteries for Nuclear Power Plants.

The batteries were mounted in a seismically approved battery rack supplied by the battery vendor ard subjected to four seismic levels, with the final level approximating the most severe that a.

Work sponsored by the U. S. Nuclear Regulatory Commission, Office of Nuclear Regulatory Research, under DOE Contract No. DE-AC07-761001570; Dr.

G.

H.

Weidenhamer, Technical Honitor.

45

l is required of batteries in the U.S.

None of the batteries experienced any loss of electrical capacity as a result of the seismic testing and only minor damage was caused of battery containers as a result of the batteries rubbing on the battery rack.

The study concluded that when batteries are maintained and operated in accordance with ILLE Std 450 and Regulatory Guide 1.129, little if any electrical capacity will be lost as a result of seismic shaking at levels that are typical of the most severe levels required for SSL in the U.S.

f The Phase I transformer study found that degradation or loss of electrical insulation is the most significant failure mode for transformers, failures of 3

the windings (turn to turn, winding to winding or winding to ground f aults) and f ailures in the bushings that provide the interf ace between the transformer and transmission lines are primarily caused by a failure of the insulation system.

The study concluded that the f ailurs rate of transformers does not show a trend that indicates an increased failure. ate with age of transformers. However over 95% of the transformers are less than 20 years old and 75% are under 15 years old.

Because transformers are normally considered to be a long life item (40 years or greater), a significant trend with age would not be expected at this time.

Because transformers in nuclear facilities are relatively young, it is recommended that a periodic review of operating experience of transformers be performed to determine whether significant trends, not previously identified, are developing.

Transformer reliability can be improved and maintained by the use of a thorough and continual program of inspection, surveillance, and maintenance.

Such a program will detect and reduce stressors that shorten transformer life, prevent stressors before they cause degradation, and detect degradation in the early stages so that preventive and corrective action can be taken prior to transformer failure.

46

Operating Experiences and Degradation Detectio;i' for Auxiliary Feedouter Systems l

l Don Casada Oak Ridge National Lateratory NRC Technical Monitor: llill Panner

SUMMARY

t The Auxiliary Feedwater (AFW) System has historically been recognited as critical to i

successful mitigation of pressurized water reactor (PWR) plant transients and accidents. A report prepared for the Nuclear Regulatory Conunission (NRC) by Oak Ridge National Laboratory (ORNL) in 1990, NUREG/CR 5404, ORNie6566/V1," Auxiliary Feedwater System Aging Study", reviewed the AISV system design, testing requirements and pracilees, and historical operating experience. This study was carried out under the auspices of the NRC's Nuclear Plant Aging Research (NPAR) program. The results of the research are presented in this paper, i

Enihire Data Redew Failure data froin the Institute of Nuclear Power Operation's Nuclear Plant Reliability Data System (NPRDS) and Licensee Event Repons were reviewed in deinil. Various parameters were evaluated during the review, including the component and subsystem affected, method of detection, and the effect of the failute on the system. The numbers of failures of i

specific components found in the databases ween not deemed to be reliable indicators of actual failure experience; however, the distrlhmlon of failures was considered to be meaningful.

Five major categories of components for the AITV system were designated for the review:

Pump drives

+

Pumps Valve operators Valves Other it was found that 37% of the system degradation was attributable to failures of pump drives, including turbines, motors, and diesels. The bulk of the drive degradation was found to occur in turbine drives, which accounted fcr 27% of overall system degradation.

. Valve power operators, including motor, air,'and electrahydraulle operators, were the

- second leading major category of system degradation, accounting for 28% of overall degradation,

• Research sponsomt by the Of 6cc of Nuclear Regulatory Research, U. S. Nuclear Regulatory Commission under Interagency Agreement DOE IM64082411 with the U, S, Department of Energy under contract No.

DE AC0$440R21400 with the Martin Marietta Energy Systems,Inc.

47 i

~, -.

e

..--r

.~,,m.--,-.r.+.-4,,.m w - m, _

,-,m-

-4+--%+-~yr.

4

,r r.,

'lhree categories of methods 01 detection were designated:

Programmatic monitoring Routine obsen ation Dentmd Failures detected during planned, periodic inspections or tests, such as suncillance tests, were deemed to have been detected by programmatic monitoring. Failures detected by operators or others dunng the course of Lonnal plant operation, such as the observation of b

valve stem leakage or a control board annunciator, were deenied to have been detected by toutine observation. Failures detected when the system was called upon to function in support of normal or emergency operations, such as failure of a pump to start on demand following a reactor trip, were deemed to have been detected on demand.

i it was found that the distribution of system degradation associated with programmatic monitoring, routine observation, and dernand was 42% 39% and 18% respectively.

l Individual Plant Design and Mepitoring Practices Review The design of a reference AIAV system for a cooperating utility's plant, and the associated surveillance, operating, and maintenance procedures were reviewed. A principal purpose of the review was to determine the extent to which potential sources of failure wo'ild be detectable by the exisiting monitoring programs.

It was found that there are two general categories of failure sources which would not have tren detected by the programmatic monitoring practices of the reference system:

Failures of various instrumentation and control components that, for whatever reason,

+

are not tested periodically, and Failures of components to perform under design basis typ0 conditions (although

+

performance under less stringent conditions may tv demonstrated periodically).

It was also found during the review of testing practices that some components appear to be tested excessively, possibly leading to accelerated aging. This was attributed to be largely i

due to the difHculty of coordincting a variety of technical specification test requirements to

- minimize testing ofindividual compoacnts (the focus of the testing program was naturally more oriented toward ensuring that all technical rpecincation sun eillances were met, not on minimizing the numbers of tests),

I CODdulions.

It was concluded that additional focus on turbine drives, specifically turbine governors and control systems, was merited. A phase I study is currently underway. It was also i

concluded that the potential for optimizing test requirements related to AFW systems should be explored. This study is scheduled to begin in FY 1992, i

1. T.

"".;~,.% '" tT m;".<o" a =;;,, 't n

=~; a::"::*

48

" O',,T f; g g

- ~ - -.,,,,

,..n-

-.,,_---..- - - - - + - n

i l

AGING ASSESSMENT OF RESIDUAh IIEAT REMOVAL SYSTEMS IN B01 LINO VATER REACTORS

• i l

Robert Lofaro Brookhaven National Laboratory i

Upton, New York 11973 l

and

~

Satish Aggarwal U.S. Nuclear Regulatory Commission War hington, DC 20555

SUMMARY

The Nuclear Plant Aging Research (NPAR) program was established by the U.S.

Nuclear Regulatory Commission to address concerns related to aging ef fects on the safety and reliability of nuclear power plants.

As part of the NPAR prograin, various studies have been performed, the goals of which are to characterize aging and service wear _ effects and identify roothods of detecting and mitigating them.

Work under the NPAR program is structured into two phases.

In phase 1, aging effects are characterized by identifying predoirinant failure causes, modes and mechanisms, along with the components most susceptible.

The second phase then uses phase 1 results to assess current surveillance and monitoring practices, and develop functional indicators _ to mitigate aging of fects.

This paper presents prelitninary phase l' results for the Residual llent Removal (RllR) system study.

In this study, the RilR system for Boiling Water Reactors was analyzed.

Varioun. designs are used, however, the teost typical includes two loops.

Each loop includes two pumps and one heat exchanger, along with numerous valves and instrumentation.

The RllR system is capable of operating in several different inode s. _ The two most. common are low Pressure Coolant Injection (LPCI) and i

Shutdown Cooling (SDC), which are the subject of this study.

The aging analysis included a review and evaluation of numerous RilR failure records from various national data base 1, as well as actual plant records.

Results showed that approxirnately.70% of the failures reported were related to aging degradation.

The dominant-failure cause was normal service, while the dominant mechanism was wear.

An evaluation of the component failures showed that va?.ve s were the component most frequently failing followed by instrumentation and controls. The valve _ failures involved mostly motor. operated valves and typically were characterized by leakage from the valve or failure of the valve to transfer. The instrumentation / control failures were predominantly switch malfunctions where the device was out of calibration or failed to' operate.

All components were found to have a-large fraction of failures related to aging.

~

- The failure records were also reviewed to determine the effect of the -

failures. At the system level it;was found that 53% of.the reported failures

• Work performed under the auspices of the U.S. Nuclear Regulatory Commission.

49

- - -, _ _. _ _----~

4

)

resulted in degraded system performance, while 21% resulted in a loss of redundancy.

At the plant level it was found that a significant number of RilR failures were the direct cause of engineered safety feature actuations, plant

{

shutdowns, automatic plant scrams, power reductions, and extension of outages.

In addition, several aging related R!lR failures were found which resulted in i

radiological releases.

This work has shown that aging degradation is a concern for PJIR systems and

- that it can adversely effect system performance, as well as plant safety.

To mitlBate the effects of aging, good functional indicators are required.

They should be capable of detecting aging degradation in the incipient stage to ensure that system reliability is not compromised. Potential functional indicators for i

the RllR system have been identified and are presented.

These results have also provided a technical basis for evaluating monitoring, inspection and maintenance practices.

A preliminary review of current practices has been made, including a survey of several utilities, and results are presented.

t t

t

-50 l

m.

wsu-q--

gd-y 9 y,.n y m*.q r

9-ysy+%

y y-.--

1r-<

ar p yvvy,=m--g4qsyww,m-ewy-m gyryeumur+y7*

5.w

-eeg-e e*.-%wd

-wa.-,ws-esr*wra--'ew4e+*== men 44-**M=tN=pW wv

l Corrosion and Erosion Effects on Valvo Friction and Operability.'

Thomas H. Hunt H. Lowell Magleby Idaho National Engineering Laboratory Dr. G. H. Weldenhamer U.S. Nuclear Regulatory Commission The Idaho National Engineering Laboratory (INEL) has studied the aging mechanisms that could affect motor-operated valves (MOVs) cperability.

These studies support the U.S. Nuclear Regulatory Commission's (NRC) Nuclear Plant Aging Research (NPAR) programs.

These studies focused on evaluating aging's effect on friction in MOVs and on the effects of valve wall thinning on MOV operability.

Our first fri: tion study reviewed corrosion, deposition, and erosion-corrosion mechanisms and their potential effect on friction, analyzed the failure data for MOVs, and inspected MOV conditions at operating facilities.

This study showed that further study of aging and friction was necessary. As a result, a second study is progressing which will corrode samples of valve materials and measure the chances in friction factors for these materials. The structur:1 analysis of wall tt inning on MOV operability is currently issued as an informal report to the NRC. This study used a finite element model of a 16-inch carbon steel globe valve to evaluate the potential for erosion-related wall thinning to compromise the rperability of these MOVs.

A draf t report, NUREG/CR 5735, " An Evaluation of the Effect of Aging on Friction Factors of Motor-0perated Valves" has been completed. The report is the product of a study that investigated the effects of three aging mechanisms, corrosion, erosion-corrosion, and deposition, on the friction coefficients of the valvn disk along the valve guides and seats of MOVs.

The study was a Phase !

NPAR Study to determine the extent of frictional related failures in MOVs and to e e.it the need for further studies in this area. Three areas were evaluated in the study: 1) a review of existing research of the three mechanisms to determine if their ef fects on friction had previously been evaluated, and if not, their likely effect on friction, 2) a review of MOV operating history to evaluate the extent of the failures due to these mechanisms, and 3) observation of the as-found condition of MOVs during maintenance activities and interviews with key M0V industry personnel to corroborate the findings of the operating history review.

' Work sponsored by the U.S. Nuclear Regulatory Commission, Office of Nuclear Regulatory Research, under DOE contract number DE-AC07-761001570; Dr. G. H.

Weidenhamer, Technical Monitor.

51

- - _ _. _ _. _ _ _ _ _l The aging-friction factor study reached the following conclusions:

The rates of corrosion of and deposition on valve sliding surfaces is not likely to obstruct the mechanical tolerances of the valves.

It is likely that corrosion and deposition products on valve sliding surfaces will cause increases in the friction factors of these i

components, but the magnitude of the problem is indeterminate requiring further study in this area.

Insufficient theoretical information is available regarding deposition on valve surfaces. The operating history review indicated a few problems related to deposition have been reported, therefore, further study is needed.

The INEL is presently conducting initial friction testing of material sampics to provide the answers to the questions raised by the previous study.

The initial friction tests will focus on carbon steel samples and carbon steel clad with Stellite. These materials are representative of the materials used in the valves of concern to the NRC's Generic Issue 87 (failure of HPCI Steam Line Without Isolation). Separate friction tests are planned on both uncorroded and

-corroded samples to assess the impact of corrosion products on sliding friction.

'The initial results of these tests are expected later this year.

The report. EGG-SSRE-10039, " An Evaluation of the Effects of Valve Body Erosion on MOV Operability" is the product of a study on erosion damage to MOVs that could compromise the valve's operability. The purpose of the study was to determine if wall thinning could reduce structural-strength so design basis stresses could deform the body preventing valve operation due to disk binding.

A finite-element model was used to simulate the effect of thinning on the structural integrity-of the valve body. The valve chosen for modeling was a 16-inch carbon steel globe valve typical of ones used in the RHR systems at BWRs.

A review of existing NRC operating history reports and sponsored research showed this system to be subject to the erosion damage of concern and it was a first-line safety' system in the event of an accident. The specific valve selection was based on the worst case observed erosion damage that has been reported by tho

=

industry.

The evaluation using the finite-element model used two sets of damage conditions: 1) i duplication of the actual damage that occurred in the worst case reported event for a first-line safety system valve, and 2) the maximum damage that could occur when discovered by normal surveillance methods.

The secor '

I conditions reduced the wall thickness of the first set of conditions to simulate through-wall erosion of the bridge and valve body.

For both of the condition i

l sets, the analysis -showed that peak stresses remained well below the yield L

stresses for the body material. For all areas of the valve body, stresses peaked at approximately 50% of yield stress.

Given these results, we concluded it is unlikely-that thinning of the valve walls-will result in deformation nf valve materials under design basis conditions.

l l

52

I EFFECT OF COMPONENT AGING ON PVR CONTROL ROD DRIVE SYSTEMS

• Edward Grove, William Gunther, and Kenneth Sullivan Brookoaven National Laboratory Upton, New York 11473 and Satish Aggarwal U.S. Nuclear Regulatory Commission Washington, DC 20555

SUMMARY

An aging assessment of PWR control rod drive (CRD) systems han been completed as part of the USNRC Nuclear Plant Aging Research (NPAR) Program.

The CRDs are flange mounted on top of the reactor pressure vene) head, and serve to position the control rod assemblies in the core in response to automatic or manual reactivity control signals.

Additionally, the systems are designed to provide a rapid insertion of the control rods upon loss of AC power. This study examined the design, construction, maintenance, and operation of the P.abcock & Wilcox (B&W), Combustion Engineering (CET, and Westinghouse (1{} systems to deterraine the potential for degradation as the system ages.

The CRD system boundaries consisted of the-drive mechanistos, pownr and control, rod position indication, and cooling system components.

The individual absorber rods, fuel assembly and upper internal guide tubes were also examined since failure of these cc.mponent s may prevent control rod insertion.

Operating experience data were evaluated to identif y the predominant f ailure modes, causes, and ef fects.

This evaluation, coupled with an assessment of the materials of construct ton and operating environment, demonstrate that each desigt is subject to degradat hn, which if l e f t.

unchecked, could af f ect '+s safety function as the plant ages.

Failures of the power and control, and rod position indication systems; and degradation of the control rod drive mechanisms and seals accounted f or the majority of system failures.

These failures have resulted in significant operational effects including primary coolant leakage, dropped or slipped rods, power reductions, plant terams, and in some lustances, emergency safety system ac tua t.lon. Since 1980, the NRC has issued six Information Notices alerting plant operators to various CRD system failures.

An industry survey, conducted with the assistance of EPRI and NUMAP.C,

identified current CRD system maintenance and inspection practices.

The results of this survey indicate that some plants have performed system modifications, replaced components, or augmented existing preventive maintenance practices in response to system aging.

The survey results also supported the operating experience data, which conc 1ried that the timely replacement f degraded components prior to f ailure was not always possible using existing condition monitoring techniques.

Therefore, the recommendations presented in this study also include a discussion of more advanced monitoring techniques, which provide trendable results capable of detecting aging.

• Work performed under the auspices of the U.S. Nuclear Regulatory Commission.

53

l AGING ASSl!SSMiiNT OF REAC~ LOR INS'IRUMENTATION AND PROTliCTION f

SYSTIIM COMPONIINTS, Pil ASl! !

A. C Gehl E W llagen Oak Ridge National 1 aboratory' W. S.1: armer Nuelcar Regulatory Commission 51ttlittW A study of the aging related operating esperiences throughout a live year period (1984-1988) of six generic instrumentation mod les (indicators, senwra, controllers, transmitters, annunelators, and recorders) was performed as a part of the USNRC Nuclear Plant Aging Research (NPAR) Program. The

~

effects of aging from operational and environmental stressors on these instrumentation nxxtules were eharacterlied using data reported in nationwide industry databases such as 1.ieensee Event Reports (LERs), Nuclear Plant Reliability Data Systems (NPRDS), and Nuclear Plant Experience (NPE). Other material used in the research included NRC Daily lleadquartern Repora, NRC Daily Operating Events Reports. NRC Regional Inspection Reports, and published literature for related investigations of instrumentation aging. Ily examination of operational history, those safety related instrumentation modules most subject to aging were identified. The data are graphically displayed as f requency of events per plant year for operating plant ages from I to 28 years to determine aging-relatel failure trend patterns.

'Ihree main conclusions were drawn from this study:

.1, Instrumentation and contiol(l&C) modules make a modest contributionto saf ety significant events.

17% of all 1.ERs issued during 1984 1988 dealt with rnalfunctions of the six I&C modules studied.

28% of the 1.ERs dealing with these I&C module malfunctions were aging related (other studies show a range of 25-50%).

2.

Of the six modules studied, indicators, sensors, and controllers account for the bulk (83%) of aging related failures.

3.

Infant mortality appears to.be the dominant aging related failure rnode for most i&C module categories (with the exception of annunciators and recorders, whleh appear to fait randomly).

Additional observations are made from examination of the module failure data, and three i

' Managed by Martm Madetta Erwrgy Systema, ino, for the ll S. Deginttnwnt or Encrry under Contred DIAC05440lCl400 54

/

recommendations were derived from the overall investigation:

1.

Consideration should be given to methods that would be letpful in reducing the incidence of infant mortality, particularly for the indleators and sensors categories, which dominate aging-related l&C module malfunctions and failures. itor example, modules could be subjected to extremes of operational environment (e g., temperature, humidity, operating voltages) and perhaps cycled over an extendeJ time period to identity marginal components prior to installation. Military standards provide good indication of those praeuces inely to be beneficial.

2.

Consideration should be given to testing selectal l&C modules for synergistic ef fects of aging stressors. The purpose would not be qualineation of specine equ;pment but rather identification and quantification of generie stress and failure relationships. Tests would not be intended to demonstrate operating envelopes of any specific brand of equipment.

1 3.

Consideration should be gis er n' sing an industry-wide databe -ledicated to aging-related information. Earlier Adies pointed out that existing databases, while reporting stressors, do not adequately indicate ti:e root cause failure mechanisms; this current study also encountered ditticulty in drawing conclusions as a result of this defi;:euey. A goad source of information for aging studies would be the maintenance records

c. the individual plants. An industry-wide, readily accessible database devoted specifically to aging-related events and information would provide a most helpful and efficient service for those interested in plant and equipment agmg, I

55 l

l

A OPERATING EXPERIENCE REVIEW OF FAILURES OF POWER OPERATED RELIEF VALVES AND BLOCK VALVES IN NUCLEAR POWER PLANTS G. A. Murphy Nuclear Operations Analysis Center OAK RIDGE NATIONAL LABORATORY W. S, Farmer NRC Technical Monitor

SUMMARY

This paper contains a review of report NUREG/CR-4692 " Operating Experience Review of Failures of Power-operated Relief Valves and Block Valves in Nuclear Power Plants" which compiled nuclear power plant operating events involving failures of power-operated relief valves (PORVs) and associated block valves (BVs). Of the 230 events identined,101 involved PORV mechanical failure,91 were attributable to PORV control failure,6 events involved design or fabrication of the PORVs, and 32 events involved BV failures. The report contains a compilation of the PORV and BV failure events, including failure cause and severity. The events are identified as to plant and valve manufacturer. An assessment of the need to upgrade PORVs and BVs to safety-grade status concludes that such action would improve PORV and BV reliability.

The greatest improvement in reliability would result from using newer, more reliable PORV designs and improving testing, diagnostics, and maintenance applied to PORVs and DVs, particularly the BV motor operator. A summary of interviews conducted with four PORV manufacturers is also included in the report.

NUREG/CR-4692 was prepared by the Nuclear Operations Analysis Center (NOAC) in response to a request from the Nuclear Regulatory Commission (NRC) Division of Engineering Technology (DET) for a survey of power-operated relief valve (PORV) and bk>ck valve (BV) operating experience. The information was provided under the Nuclear Plant Aging Research -

(NPAR) Program to support the resolution of Generic Issue 70 (GI 70)"PORY and Block Valve

' Reliability."

'Research sponsored by the Of0cc of Nuclear Regulatory Research, 0, S. Nuclear Regulatory _

Commission under Interagency Agreement DOE 1886-8082-8B with the U. S. Department of Energy under contract No. DE-AC05-840R21400 with the Martin Marietta Energy Systems, Inc.

.1wi"".",,".~ffs % ITs 1T ;&a 2% *L &

' 0 ".". L I. ~. :. a ~ ~ " "

56 L'""**,,'";"'," Ogg-M 4

- - -. _. -. -. ~ - -. - - -. _

y 9

~ ASSESSMENT OF DIAONOSTIC METilODS FOR SOLENOID. OPERATED VALVES

-Robert C. Kryter -

Oak Ridge National Laboratory W. S. Farmer Nuclear Regulatory Commission

&mmarv Solenoid-operated valves (SOVs) were studied at Oak Ridge National Laboratory as part of the

. USNRC Nuclear Plant Aging Research (NPAR) Program. The primary objective of the study was to: identify, evaluate, and recommend methods for inspection, surveillance, monitoring. and maintenance of SOVs that can help ensure their operational readiness-that is, their ability to perform required safety functions under all anticipated operating conditions.

Solenoid-operated valves are available, both with and without nuclear qualification, from a number of different rnanu eturers and are found throughout nuclear power plants in relatively large numbers, r

oftentim:

>. component oflarger, more complex, and clearly safety.related systems such as containsny m

've actuators, BWR control rod scram systems, and PWR safety injection lly simple devices, with a long history of satisfactory operation in a variety syste

• Thev s

of-t n-nuclear industrial app!! cations. However, their presence in systems 4

impc it '

res an especially high degree of assurance that they are ready to perform their n ader all anticipated operating conditions, since failure of one of these small andrd

< <c devices could have serious consequences under certain circumstances.

An earlier (Phase 1) NPAR program study described SOV failure modes and caeses and identified measurable parameters thought to be linked to the progression. of ever-present degradation mechanisms that may ultimately result in functional failure of the valve, Using this earlier work as a guide, the present (Phase II) study focused on devising and then demonstrating the effectiveness

-of techniques and equipment with which to measure performance parameters that show promise for detecting the presence and trending the pregress of such degradations before they reach a critical stage.

Intrusive techniques requiring the addition 'of magnetic or acoustic sensors or the application 01

. special test signals were investigated briefly, but major emphasis was placed on the examination of -

condition indicating techniques that can be applied with minimal cost and impact on plant operation -

_ (see accompanying Table it These include remotely monitoring coil mean temperature, determining

~

valve plunger position and verifying unrestricted SOV plungd movement, and detecting the presence of shorted turns or insulation breakdown within the solenoid coil. The first of these techniques, though perhaps the simplest conceptually, will likely benefit the nuclear industry most because SOVs have a history of failure in service as a result of unwitting operation at excessive temperatures.

Experimental results are presented that demonstrate the technical feasibility and practicality of the monitoring techniques assessed in the study, and recommendations for further work are provided.

L

!-o

' Managed by Martin Marietta Energy Systems. Inc,. ror the U.S Department of Energy under Contract DE-AC054 tor 21400.

57

~

_ ~

__m t

Table 1. SOV monitOfing methods evaluated in this study Degradation (s) Or malfunction (s)

Promise for in.

Method addressed Attributes plant use Nonperturbathe to plant operations liigh; ceady for Measurement of SOY Electrical failure of coil e

ten,peratute, via coil and degradation of No new senson or signal cables are immediate uw e

resistance or impedance clastomers resulting from required prolonged operation at No permanent instnamentation e

cxcesshely high required; can be applied as needed temperatures from a remote location Applicable to ae. and de-powered SOVs No need for add on sensors or signal 11 gh; some addiuonal Indication of valve position Mechanical binding, e

and clunge of state upon slugghhness, or failure to cables development work Valve position readout from a remote required t.pphcation of power, via shift as a result of worn or e

change in coil impedance improper parts or the location e Static method does not disturS SOV presencv of foreign materials inside the valve Detects simultaneously 4*egradction Medmmi further Indicatk n of mechanical Mechsencal binding and e

bindmg. t'y tracking changes - sluggish shifting caused ty of magnetic or spring to rea, an I testing needed to in current and voltage at worn,.wollen, or improper increase in frictional forxs ascertain cause of e No need for add on sensors or cables - poor repeatability of SOV pull-in and drop +ut -

parts or the presence of points foreign materials inside the or access to SOV test results Applicable to ac. and de-powered valve e

SOVs Detects presence of defects within I.ow; useful for Indication of shorted coil Electrical failure of e

turns or insulation solenoid coil, caused by coil that cannot be revealed bv other lateratory post.

breakdown, based on high-voltage turn off means monem tests characteristics of electrical transients in combination transient generated upt n with insulation weakened deenergizing a de SOV by prolonged operation at high temperatures No need for adden sensors, signal Minimal; indication of mechanical Mechanical binding and binding, by analyzing the sluggish shifting caused by canes, or access to SOV

. investigation of i

time-varying characteristics of worn, swollen, or improper e Information could be obtained as a method abandoned

- the inrush current parts or the presence of -

result of ewryday valve operation.

early in the study accompanying application of foreign materials inside the electrical power to the SOV vahc Indication of mechankal -

-' Wear of imernal valve,

No need for adden sensors, signal Minimal; looseness within ae-powered parts, improper assembly,.

cables, or access to SOV hwestigation of Nonperturbative to plant operations method abandoned valves, via electrical detection - or replacement with of humming or chattering of incorrect parts early in the stuoy.

the plunger assembly Additioti of miniature (frequency decomposition of acoustic sensor to steady-state coil current)

SOV might prove worthwhile c

58

SHIPPINGPORT STATION AGING MANAGEMENT LESSONS R. P. Allen Pacific Northwest Laboratory J. J. Burns U.S. Nuclear Regulatory Commission The decommissioned Shippingport Atomic Power Station has been a major source of naturally aged equipment for the U.S. Nuclear Regulatory Commission's Nuclear Plant Aging Research Program and other aging research studies.

The-decommissioning of the Shippingport Station provided a unique opportunity to conduct in situ assessments at an aged reactor and to obtain a variety of naturally aged components and samples for detailed laboratory evaluation.

As the first U.S. large-scale, central-station nuclear plant, the Shippingport Station paralleled commercial pressurized-water reactors in reactor,-steam, auxiliary, support, and safety systems.

Its 25-year service life (1957 to 1982) overlapped the construction and initial operating period of most reactors currently operating. Also,-because of su tantial modifi-

)

cations during the mid-1960s and 1970s, the Shippingport Station offered unique examp'ies of identical or similar equipment used side-by-side, but representing different vintages and degrees of aging.

As part of the Pacific Northwest Laboratory (*) Shippingport Station aging evaluation work, more than 200 items,-ranging in size from small instruments and materials samples to one of the main coolant pumps, were removed and shipped to various laboratories for evaluation.

These items included battery chargers, inverters, relays, breakers, switches, power and control cables,- electrical penetretions, check valves, solenoid valves, and motor-operated valves.

Samples of piping from various plant systems also were acquired for radiological characterization studies, and samples from the primary system check valves, main stop valves, ar.d main coolant pumps.were removed for materials degrudation studies.

In situ assessments of Shipping-port Station components also were conducted, including the pre-removal visual and physical examination of-components, the testing of electrical-circuits, and special measurements to assist in selecting specific components for further evaluation.

Naturally aged components and materials are subjected to the actual in-plant environments, operating conditions, testing procedures, and mainte-nance practices. Thus,- their evaluation is an important way to verify expected degradation mechanisms and failure modes, to ensure that other l

components will be minimally affected by aging, to validate aging projections

-based on the extrapolation of accelerated test data, and to detect unexpected (a)

Pacific Northwest Laboratory is operated for the U.S. Department of l

Energy under Contract DE-AC06-76RLO 1830.

Work is conducted under NRC l

FIN 82911.

?

i 59 1

~..

aging mechanisms (surprises) that could significantly impact component or system performance. The following are examples of the types of aging manage-ment information that have been derived from the studies of the naturally aged Shippingport_ Station components and materials.-

Identification of Eouioment Demonstratina Satisfactor.y lona-term Performance with Minimal Aoina Effecti Naturally aged inverters and battery chargers from the Shippingport Station were tested by Brookhaven National Laboratory. Although some aging induced changes were noted, it was concluded that aging had not sub-stantially affected equipment operation.

Similarly, Wyle Laboratories conducted operability and performance tests of Shippingport Station circuit breakers and relays, demonstrating that these components still met the original specifications.

fomoarison Data for Validatina Aoina Projections Based on Accelerated Aaing Studies An Argonne National Laboratory investigation of the microstructural characteristics of cast stainless steel from Shippingport Station primary system valves _ and pump volutes verified that the low-temperature thermal embrittlement._ mechanism for this naturally aged material is the same as that of-laboratory-aged material. This provided a unique opportunity to validate and benchmark the laboratory studies.

Basic-Insiahts on Dearadation Mechanisms An Argonne National Laboratory evaluation of samples from the inner wall of the Shippingport Station neutron shield tank, which represent base metal and weld material exposed to different neutron flux levels, has provided valuable information on possible low-temperature low-flux embrittlement processes in reactor pressure vessel support structures.

Detection-of Unexoected Aoina Mechanisms (Surprises)

An Oak Ridge National. Laboratory evaluation of a piston lift check valve

.from the Shippingport Station found significantly more wear than would be expected based on the valve's normal service environment.

Sirallarly, an evalcation of an 8-in, diameter gate valve and operator from Shippingport Station revealed a previously unrecognized cable sizing _ problem.

identification of Condition and Performance Parameters to Detect and Monitor Aoina An Idaho National Engineering Laboratory in-situ evaluation of Shippingport Station electrical circuits confirmed the effectiveness of the measurement system for detecting degradation of circuit connections and splices.

60

Effectiventes of Surveillanco Methods for the Class 1E Power and Reactor Protection Systems' V. Sharma Idaho National Engineering Laboratory The Idaho National Engineering Laboratory (INEL) performed aging studies of the Class IE Power (IE) and Reactor Protection Systems (RPS) in support of the United States Nuclear Regulatory Commission's (USNRC) Nuclear Plant Aging Research (NPAR) Program.

The Phase I research results have been presented in past reports and papers.

This Phase 11 research program consisted of: 1) an examination of operational data to identify those failing components that were not being detected by routine (planned) inspection, surveillance and monitoring methods (IS&MM), with the respective failures causes; 2) an ideitification of the risk significant components in the Class IE Power system and;theeffectsofagingupontheirfailureprobabilities;and3)an inv~stigation of advanced and improved IS&MM practices applicable to the Class lE h d Reactor Protection Systems.

f The operational data review showed that approximately 50% of all Class lE Jawer System failures were not detected by routine IS&MM.

Batteries, circuit breakers, inverters, and relays accounted for over 80% of all IE System failures.

Furthermore, these components dominate the failures in each of the four subsystems in the Class IE Power system; the Plant AC (PAC), the Instrument AC (IAC), the DC Power (DCP), and the Emergency Power (EMP) subsystems. This indicates that the performance of a subsystem can be significantly improved by concentrating on the detection methods of just one or two types of components that are significant for that respective subsystem, m

The data showed that, for the DCP subsystem, batteries had the largest number of failures.

Routine IS&MM for these batteries were not completely effectivt in detecting failures caused by short/ ground, open circuit, burned circuit, wear, and aging / cyclic fatigue, in the PAC subsystem circuit breakers had the highest numbers of failures; its routine methods did not effectively detect failures caused by open circuits, weld faults, out-of-adjustment, and mechanical damage.

Inverters had the largest number of failures in the IAC subsystem; short/ ground circuit and open circuit failures were not effectively detected by the routine methods.

For EMP subsystem relays and circuit breakers had the highest number of failures.

Routine IS&MM was not effective in detecting open circuit, burned circuit, out-of-adjustment, and wear relay

a. Work sponsored by the U. S. Nuclear Regulatory Commission, Office of Nuclear Regulatory Research, under DOE Contract No. DE-AC07-761001570; Dr. G. H. Weidenhamer, Technical Monitor.

61

failures. Open circuit, wear, burned circuit, and out-of-adjustment failures were not effectively detected by routine IS&MM for the circuit breakers.

The operational data review also revealed that approximately 40% of all Reactor _ Protection System failures were not detected by routine IS&MM. Over 70% of all_RPS_ events were accountable to transmitters, integrators, and bistables.

For transmitters routine methods were not effectively able to detect failures caused by wear and aging / cyclic fatigue.

Open and dirty circuit failures were not detected effectively by routine IS&MM for bistables.

Aging / cyclic fatigue and defective connections caused the failures not effectively detected by these routine methods for integrators.

A standard Probabalistic Risk Assessment (PRA) was used to determine the risk significant Class IE components. Truncation of the fault tree cut sets left the circuit breakers, transformers and buses as the risk significant components. The effects of aging were determined by simply applying aging rates, as determined by expert opinion, to the initial failure rates used in the PRA to determine if significant changes occurred as a function of time.

The failure rate for the circuit breaker increased the most at 20%, followed by the transformer (6%) and the bus work (1%).

Requantification of the PRA, using the new failure rates, revealed a less than 1% increase in the core damage frequency.

This was verified by using operational data and statistical methods to formulate an-actual aging failure rate for the circuit breakers, since they had the highest theoretical failure rate increase, An in-plant evaluation was performed to study ways to improve current

-lS&MM practices for both the Class IE and Reactor Protection Systems, The cooperating plant currently uses a variety of advanced IS&MM.

One method-uses electromagnetic signals to measures various parametert of de-energized electrichl circuits.

This system determines insulation integrity of the system through measurements of capacitance, dissipation factor, impedance, and insulation resistance.

Time Domain Reflectometry is used to measure circuit integrity.

This method collects baseline data and uses it in a comparison on later-tests of the same piece of equipment; this has proven especially effective for circuits. Other improved methods include the-use of state-of-the-art infrared technology. This method.can be used on a "one time only" basis to reveal hot spots and to collect data for long term trending.

-The basic. principle of-this device is to measure the heat generated by-electrical _ equipment; the test data is compared to baseline criteria to determine whether that component is within a certain operating range.

Data obtained from this system is stored on computers for analysis. A third system used at this plant is Redundant' Instrumentation Monitoring.

This system is used_to verify calibration of on-line-instrumentation.

A computer reads instrument-output data-from the_ plant computers. The system then allows comparison of redundant channel outputs over time or trending of one channel.

The ability of all these systems to store data on computers (in an easily retrievable form) allows plant personnel the ability to-trend the IE Power and s

Reactor Protection Systems for aging related degradation.

l-62

EFFECTIVE AGING MANAGEMENT OF CIRCUIT BREAKERS AND RELAYS by i

James F.

Gleason, P.E.

Wyle Laboratories It is the challenge of a good preventive maintenance program to be sensitive to the effects of aging.

Early identification of age related degradation increases the probability that the safety significance of this aging is minimized.

An effective inspection, surveillance and monitoring program enhances mitigation of the h

impact of age related degradation on the safety of nuclear plant operations.

This paper describes the results of a comprehensive aging assessment of relays and circuit breakers that was completed as part of the NRC Nuclear plant Aging Research (NPAR) Program. Relays and circuit breakers were analyzed because they are important safety-related equipment which perform critical functions in the operation and control of nuclear power plants and have experienced age related degradation.

The research investigated current and improved inspection, surveillance and monitoring methods (ISM) for circuit breakers and relays.

The effort was accomplished in four major elements. These were an investigation into current and advanced ISM motheds, tests of aged relays and circuit breakers, tests of degraded.elays and circuit breakers and in-situ tests.

The results are significant in that :

o Improved inspection, surveillance and monitoring (ISM) methods for relays and circuit breakers have been identified which are more effective at detecting aging degradativt than current nuclear plant practice, o

Less intrusive ISM methods, which have the potential for providing predictive maintenance and condition based maintenance have been determined, and Important modifications of specific maintenance practices o

for some common relays and circuit breakers are presented where the research showed superior and more cost effective ISM methods for

-these devices.

63

Current and advanced ISM methods were ascertained by soliciting information from nuclear and non-nuclear utilities, relay and circuit breaker manufacturers and maintenance facilities.

Testing of aged devices was performed. Test specimens fcr each of the ?ive relay (auxiliary, control, electronic, protective and timing) and two circuit breaker (molded case and metal clad) types were solicited from nuclear and non-nuclear utilities and manufacturers. A total of 39 specimens were tested to the current and improved ISM methods.

Eleven specimens of relays and circuit breakers were purposely degraded and the ISM methods performed after each degraded condition. The purpose of these degradation tests was to evaluate the effectiveness of the method to detect and/ or predict the level of degradation. This also provided some quantifiable parameters of the extent of degradation. The degradations chosen for each relay and circuit breaker type were purposely severe, but for the most part, did not cause total loss of operability of the device. Thus, it was an attempt to simulate the worst state of deterioration or degradation prior to failure to operate.

The degradations were chosen based on a review and evaluation of the failure mode and mechanisms reported in NPAR reports, specified by the nuclear and non-nuclear utilities, manufacturers, and experiences of the research team.

9 These degradation condition evaluations showed that generally accepted current nuclear maintenance practices do not always detect significant aging mechanisms.

This result provides insight into the reason why failures of safety-related relays have occurred in service in spite of a comprehensive maintenance program.

The practicability of the effective methods was also evaluated at Duke Power Company's Catawba Nuclear Station and Niagara Mohawk Power Corporation's Nine Mile Point Unit 1 Nuclear Plant.

While at the

plants, the research team witnessed plant maintenance personnel performing routine plant maintenance on relays and circuit breakers.

Copies of procedures were obtained, results of plant maintenance tests were reviewed, and engineering and maintenance personnel were interviewed.

Additionally, non-intrusive ISM methods of infrared pyrometry, infrared scanning and vibration testing were demonstrated to the plant personnel.

The plant maintenance personnel and engineering staff at both plants were found to be cooperative, professional, experienced, knowledgeable and eager to discuss potential improved techniques.

64

UNDERSTANDING AND MANAGING EFFECTS OF BATTERY CHARCER AND INVERTER AGING

• William Gunther Brookhaven National Laboratory Upton, New York 11973 and Satish Aggarwal U.S. Nuclear Regulatory Commission Washington, DC 20555

SUMMARY

An aging assessment of battery chargers and inverters was conducted under the auspices of the NRC, Nuclear Plant Aging Research (NPAR) Program.

The intentions of this program are to resolve issues related to the aging and service wear of equipment and systems at operating reactor facilities and to assess their impact on safety, inverters and battery chargers are used in nuclear power plants to perform significant functions related to plant safety and availability.

The specific impact of a battery charger or inverter failure varies with plant configuration.

Operating experience data have demonstrated that reactor trips, safety injection system actuations, and inoperable emergency core cooling systems have resulted from inverter failures; and de bus degradation leading to diesel generator inoperability or loss of control room annunciation and indication have resulted from battery and battery charger failures. For the battery charger and inverter, the aging and service wear of subcomponents have contributed significantly to equipment failures.

To identify aging and service wear effects and appropriate inspection / surveillance / monitoring techniques, it was necessary to examine potential failure modes, mechanisms, and causes, This was achieved by reviewing battery charger and inverter design and materials of construction, by establishing the stressors that are both operational and accident related, and by reviewing existing failure related data.

The primary contributors to inverter and battery charger failures are overheating, electrical transients, and personnel errors.

In many cases, the stresses induced from these occurrences results in an accelerated aging of critical components.

Electrolytic capacitors, fuses, magnetics (inductors and

. transformers) and semiconductors such as silicon controlled rectifiers, are susceptible to aging degradation resulting from these stresses.

Based on the testing of a naturally aged inverter and battery charger, surveys of the current maintenance practices at nuclear power plants, and an assessment of the present equipment technology, it was found that methods exist for detecting aging degradation.

The detection techniques found to be most valuable in determining inverter and battery charger aging are component and

• Work performed under the auspices of the U.S. Nuclear Regulatory Commission.

65

_ m. _ __ _ _.

equipment temperature monitoring, periodic observation of circuit waveforms, and component parameter measurements, Mitigating the effects of aging degradation can be accomplished through a combination of maintenance, design changes, and personnel training.

It is recommended that a comprehensive maintenance program be established for inverters and battery chargers that ' addresses four areas: -inspection, testing, predictive maintenance, and corrective maintenance. Guidelines for each of these areas have been provided in research reports NUREC/CR-4564, 5051, and 5192.

Current maintenance practices at nuclear power plants were evaluated to assess utility programs for addressing inverter and charger aging.

Two independent surveys,.one by BNL and one by the Electric Power Research Institute (EPRI)- found - that all of the surveyed plants specified some maintenance activities for chargers and inverters, and most utilities are cognizant of this equipment's importance to safety and availability.

However, the wide range in the type of maintenance performed, and the varying levels of success achieved could reflect inadequacies at some plants.

While emphasis has been placed on maintenance practices for mitigating equipment aging, the potential improvement in vital bus reliability through design improvements should not be overlooked.

As plants age, there should be an awareness of improvements in equipment to take advantage of advances which could minimize the plant's aging effects and maintain or enhance. its established performance and safety goals. One recommended design improvement is 'the autoestic transfer switch. The device reduces the impact of inverter degradation by sensing the failure and switching the vital bus to an

- alternate ' electrical source without. interrupting power to safety related instrumentation, controls,-and equipment.

Other recommended improvements resulting from this research include the application of equipment for detecting and suppressing electrical bus transients experienced regularly in power plants, the use of higher voltage and temperature rated components in the inverter circuitry, and the addition of forced air

- cooling to reduce the overheating problems experienced.

With personnel induced stresses accounting for approximately 15% of the age related inverter and charger - failures, it is recommended that training be l

provided and procedures be established for the operation and maintenance of this L

fairly complex equipment.

Maintenance must be performed periodically to refurbish and/or replace components' 'which exhibit aging.

In addition to discrete components such as capacitars, transformers, and semiconductors, the integrity of other entities such as cable-connectors,- wiring, and structural fasteners _ must also be maintained -- to assure proper ; equipment operation under normal operating and l1 postulated accident conditions.

While it is not possible to prevent all L

component failures, preventive maintenance activities and condition monitoring-L techniques can be effective in reducing the number of failures.

66 L

i n

--s n

-.w->

SAND 92 0206A i

Aging Assessment of Cables

• Mark J. Jacobus Sandia National laboratories Albuquerque, NM 87185 Abstract This paper summarizes the results and conclusions of aging, condition monitoring, and accident testing of Class IE cables used m nuclear power generating stations. The primary objectives of this experimental program were to determine the life extension potuual of popular cable products used in nuclear 30wer plants and to determine the potential of condition monitoring (CM) for residua life assessment. Cable insulation materials that were tested include cross-linked polyethylene (XLPE), ethylene propylene rubber (EPR), polyimide, and silicone rubber (SR). Cable jacket materials that were tested include neoprene, chlorosulfonated polyethylene, and fiberglass braid.

Three sets of cables were aged using simultaneous thermal (-100 C) and radiation

(-0.10 kGy/hr) conditions. One set of cables was aged for 3 months, one set was aged for 6 months, and one set was aged for 9 months. After aging, each set of cables was

. exposed to a simulated loss-of-coolant accident (LOCA) consisting of high dose rate irradiation (-6 kGy/hr) followed by a high temperature steam exposure. A fourth set of cables, which were unaged, was also exposed to the LOCA conditions. The cables that were aged for 3 months and then LOCA tested were subsequently exposed to a high temperature steam fragility test (up to 400 C), while the cables that were aged for 6 months and then LOCA tested were subsequently exposed to a 1000-hour submergence test in a chemical solution.

The experimental program provided several improvements upon most previous efforts by employing considerably less accelerated, simultaneous thermal and radiation aging conditions; by employing many more condition monitoring measurements during agm, g; by performing similar accident tests on cables aged to four differem nominal hfetimes (including unaged cables); by obtaining data on t1e submergence behavior of cables that had been exposed to aging and LOCA testing; and by providing information on the ultimate thermal fragility of many different cable products after exposure to aging and LOCA testing. The test program generally followed the guidance of IEEE 323-1974 and IEEE 3831974.

The accelerated aging temperature was determined by equating the 6-month exposure to a 40-year life and assuming an activation energy of 1.15 eV and a plant ambient temperature of 55 C. The accelerated radiation aging dose rate was determined by assuming a 40-year radiation dose of 400 kGy. The 3-month chamber was therefore nominally equivalent to 20 years of aging and the 9 month chamber was nominally equivalent to 60 years of aging.

• The-Aging Degradation of Cables Program is supported by the United States Nuclear Regulatory Commission and performed at Sandia National Laboratories, which is operated for the U.S. Department of Energy under contract number DE AC%76DP00789.

67

During the aging exposure, various electrical and mechanical condition monitoring measurements were performed on the cables. Some of the measurements were performed on completed cable samples and others were performed on smaller samples that were aged together with the comple'te cable specimens and were removed from the test chambers during aging. The parameters measured included insulation resistance and polarization index at three different voltages, capacitance and dissipation factor over a wide range of frecuencies, clongation and tensile strength at failure, modulus profiles, cable indenter modulus tests (using a cable indenter developed at Franklin Research Center under Electric Power Research Institute funding), hardness, and bulk density.

The results of the tests indicate that the feasibility of life extension of many popular nuclear power plant cable products is promising and that mechanical measurements (primarily elongation, modulus, and density) are more effective than electrical measurements for monitoring age-related degradation. In the high temperature steam test, ethylene aropylene robber (EPR) cable materials generally survived to higher temperatures 11an crosslinked polvolefin (XI.PO) cable materials, in dielectric testing after the submergence testing, th'e XLPO materials performed better *han the EPil materials.

The aaper will summarize the preliminary conclusions from the overall test program, divit ed into six categories: general conclusions regarding aging and condition monitoring, general conclusions regarding the accident performance of aged cables, specific conclusions from accident testing of each of the three specific cable types (XLPO, EPR, and miscellaneous), and general conclusions from the submergence and high temperature steam tests.

68

ASSESSMENT OF DIAGNOSTIC METHODS FOR DETERMINING DEGRADATION OF CHECK VALVES

• H.

D. Haynes Oak Ridge National Laboratory W.

S.

Farmer U.

S. Nuclear Regulatory Commission

SUMMARY

Check valves are used extensively in nuclear plant safety systems and balance-of-plant (BOE, systems. Their failures have resulted in significant maintenance efforts a..d, on occasion, damage to other flow system components.

~

Check valve failures have largely been attributed to severe degradation or internal parts resulting f rom instability (flutter) of chr ;' valves under normal plant operatir7 conditions.

The Oak Ridge National Laboratory (ORNL) has carried out a comprehensive aging assessment of check valves in support of the NRC Nuclear Plant Aging Research

', W PAR ) program. This paper provides a summary of that assessment with emphasis on the identification, evaluation, and application of check valve monitoring methods and techniques.

As part of the ORNL Advanced Diagnostic Engineering Research and Development Center, ORNL has developed two novel nonintrusive methods that are useful for monitoring the position and motion of check valve internal parts. These methods are based on the use of externally-applied magnetic fields from permanent magnets and from electromagnet coils driven by either alternating or direct current.

Descriptions and evaluations of several check valve monitoring methods are provided in this paper including:

a Acoustic emission monitoring a Ultrasonic inspection a Magnetic flux signature analysis (MFSA) u External ac-and de-magnetic techniques A major conclusion reached was that none of these methods examined could, by themselves, monitor the position and motion of valve internals and valve leakage; however, the combination of acoustic emission with either af the other methodo yields a monitoring system that succeeds in providing the means to determine vital check valve operational information.

Other areas covered in the paper include descriptions of relevant regulatory issues and activities, other related diagnostica resca h at ORNL, and interactions ORNL has had with ' outside organizations for the purpose of disseminating research results.

Research sponsored by the Office of Nuclear Regulatory Research, U.S.

Nuclear Regulatory Commission under Interagency Agreement DOE 1886-8082-8B with the U.S. Department of Energy under contract No. DE-AC05-840R21400 with Martin Marietta Energy Systems Inc.

w

.ac a

m, c.

r nn

.,e-.a av. coresca,cas e u a c-w.

certe.st ho

• Acme 40R2i400 Accarene, s ua 69 c.-,-..

r. un e

th. Os# Ashed $tstet C# St.

^ M Of adoew OU*.rt to du ec Ord u.a. GOugrnment papeas "

AN O'ERVIEW OF TIE STRUCTUPAL AGING PROGPM Dan L Naus Oak Ridge National laboratory (ORNL)

Oak Ridge. 'IU 37831-8056 E. Gunter Arndt E U.- S. Nuclear Regulatory Ccanission (NRC)

Washington, D. C. 20555 ABSTRACT The Structural Aging (SAG) Program was initiated at OPNL in mid-1988 and has the overall objective of preparing a report khich will provide NPr license reviewers and licensees with the following:

(1) identification

-and evaluation of the structural degradation processes: (2) issues to be addressed:under nuclear power plant continued service reviews. as well as criteria, and their bases, for resolution of these issues:

(3)'

identification and evalu : tion of relevant inservice inspection or structural assessment programs in use, or needed; and (4) - quantitative methodologies - for assessing.

current, or predicting

future, structural safety margins.

The SAG Program consists of three technical tasks:

MaterialsL -Property Data-Base.

Structural Component a

Assessment / Repair Technology.-

and Quantitative Methodology for Continued Service Determinations.

The objective of the materials property data base task is-to develop a computer-based structural materials property data base which will

' contain ~information on the time variation of material properties under the influence of pertinent environmental stressors and aging factors.

Do ccrnplementary data base formats, hardcopy and electronic, have been

' developed and: : contain inforTnation.'on, the performance. of concrete, conventional steel reinforcement.- prestressing. steel, and ' structural

> steel materials. ' Also :under this task a state-of-the-art report has

-been prepared which -identifies and - evaluate's models -. and accelerated-

aging. techniques and methodologies which can be -used in making

- predictions.of the remaining service life of - concrete in nuclear power fplants.

• (.

Pascarch sponsored by the Office of Ibclear Regalatory Research. U.S.

Naclear Regalatory Com

ission under interagency Agrecrent 1886-8084-5B with the U.S.

Department of Energy under Contract DE AC05-840P21400 with Partin Parietta Energy E Systms.'Inc.

The submitted nanascript has been authored by a contractor of the U.S.

Goverrurent under Centract lb. DE-AC05-840P21400. Accordingly, the U.S. Goverrment

-retains a ncrr.xclusive, royalty-free license to publish cr reproduce the published

. *crm of this contribution, or allos others to do so for U.S. Government purposes.

-i 70

The overall objectives of the structural emponent assess mnt/ repair technology task are to (1) develop a systmatic nethodology which can be used to make a qt4antitative assescrent of the presence, tragnitude, and significance of any enviromental sitessors or aging factors, (2) provide recanended inservice inspection or sa::pling ptocedures, and (3) identify and evaluate techniques for mitigation of any envirotrental stressors or aging factors which may act on critical concrete cmponent s.

Under this task an aging assencrent nethodology has been developed which can be used to identify and rank concrete structures as well as the degradation factors which can inpact the perfonrance of these structures.

Also, a s t at e-o f - the-art report has been prepared which reviewed and assessed inservice inspection techniques and methodologies for application to concrete structures in nuclear pcwer plants.

Tne overall objective of the quantitative nethodology for continued service determinations task is to develop procedures that can be used to perfarm condition assescients and make reliability based life predictions of critical concrete cmponents in nuclear power plants.

A draft report has been empleted which describes developtent of a orobabilistic nothod for condition assessnent and life prediction of concrete structures.

The rescits of the SAG Program will provide an inproved basis for the NRC stafi to evaluate requests for continued operation beyond the ncritinal 40-year design life of a nuclear power plant.

Potential regulatory applications of this research include:

(1) improved predictions of long-term material and structural perfonnance and available safety margins at future tines, (2) establistrent of limits on exposure to environmental stressors. (3) reduction in total reliance by licensing on inspection and surveillance through development of a methodology which will enable the integrity of sttuctures to be assessed (either pre-or post-accident),

and (4) inprovenents in damage inspection methodology through potential incorporation of results into national standards which could be referenced by standard review plans.

71

l DATA B ASE ON STRUCTURAL hiATERIALS AGING PROPERTIES

• C. B. Oland Oak Ridge National Laboratory Oak Ridge, Tennessee ABSTRACT The U.S. Nuclear Regulatory Commission (USNRC) has initiated a Structural Aging Program at the Oak Ridge National Laboratory to identify potential structural safety issues related to continued service of nuclear power plants and to establish criteria for evaluating and resolving these issues. One of the tasks in this program focuses on the establishment of a Structural hinterials Information Center (SMIC) where data and infom1ation on concretes and other related materials used in nuclear power plant construction are being collected and assembled into a struc-tural materials property data base. This data base will be used to establish current properties for materials in existing concrete structures and to predict future perfonnance of these materials.

The Structural hiaterials Infonnation Center is being presented in two complementary for-mats. The Structural Materials Handbook is published in four volumes as an expandable, hard-copy, reference document. The Structural Materials Electronic Data Base is formatted for use on an IBM-compatible personal computer. The SMIC contains baseline data, reference properties, and environmental information that are presented as tables, notes and graphs, The handbook con-tains a complete set of data and infonnation for each material included in the data base and serves as the reference infomiation source for the electronic data base. The electronic data base enhances the use of the handbook by providing an efficient means for searching the various data base files to locate materials with similar characteristics. The handbook and the electronic data base will be used by USNRC reviewers who perform structural assessments for continued service.

Material properties, data, and information are being collected at the Structural Materials Information Center from open literature and published references, and from laboratory tests con-ducted on material samples removed from existing concrete stmetures. Currently, the data base includes properties for portland cement concrete, metallic reinforcement, prestressing tendon and structural steel materials. As data and infonnation for other materials such as rubbers, plastics, and nonferrous metals are obtained, the data base will be expanded and updated.

• Research sponsored by the Office of Nuclear Regulatory Research, U.S. Nuclear Regulatory Commission under Interagency Agreement 1886-8084 5B v ith the U.S. Department of Energy under Contract DE-AC05 840R21400 with Martin Marietta Energy Systems, lac.

The submitted manuscript has been authored by a contractor of the U.S. Govemment under Contract No. DE-AC05-840R21400. Accordingly, the U.S. Govemment retains a nonexclusive, royalty-free license to publish or reproduce the published form of this contribution, or allow others to do so, for U.S. Government purposes.

72

Reliability-based Condition Assessment of Concrete Structures in Nuclear Power Plants Bruce Ellingwood and Yasuhiro Mori Department of Civil Engineering lohns Hopkins University Baltimore, Maryland 21218-2699 Introduction During the next 15 years, the operating licenses for a number of nuclear power plants in the United States will expire.

Faced with the prospect of having to replace the lost generating capacity from other sources and substantial shutdown and decommissioning costs, many utilities are expected to seek renewals of their plant operating licenses.

Reinforced or prestressed concrete structures provide essential safety-related functions in a nuclear plant.

Studies have shown that replacement of or major repairs to certain concrete structures in the plant as a condition of continued service would be economically unfeasible.

Concrete structures may be affected by aging, or changes in strength and stiffness beyond the baseline conditions assumed in initial structural design.

Some of these aging effects are benign; others may cause strength to degrade over extended periods of time, particularly when the concrete is exposed to an aggressive environment.

Thus, the evaluation of safety-related con; rete structures for continued service should provide quantitative evidence that their strength will be sufficient to withstand future extrere events within the proposed service period wilh a level of reliability sufficient for public safety.

Technical Acoroach A methodology is being developed as part of the Structural Aging Program to evaluate the time dependent reliability of a reinforced or prestressed concrete structure.

This methodology takes into account the stochastic nature of past and future loads due to plant operating conditions and the environment, and randomness in those physical processes and environmental stressors that may lead to degradation in strength.

The role of periodic inspection and maintenance in meeting a target reliability level over a period of continued service is also included.

Events giving rise to significant structural loads occur randomly in time. When viewed on a timescale of 40 to 100 years, however, such events occur infrequently, have a relatively short duration, and occupy 73

only a small fraction of the total life of a structure.

Based on these observations, many of the operating, anvironmental and accidental loads that act on nuclear plant structures can be modeled as Poisson renewal or pulse processes.

Statistical data to describe the rates of occurrence, duration and intensity of these load events are being synthesized from other research programs.

Changes in engineering properties of steel and concrete over an extended service life are modeled as time-dependent random functions.

Strength degradation mechanisms related to corrosion of reinforcement, detensioning of prestressing tendons, and aggressive chemical attack from sulfates and acids appear potentially important for concrete structures in nuclear power piants.

Statistical data to describe initial strength are available from previous probability-based code studies; strength changes in time are based on data provided from another task in the Structural Aging Program.

Structural reliability analysis integrates the probabilistic descriptions of strength and loads to describe time-dependent reliability and deterioration of concrete structural components and systems subjected to stochastic loads. The reliability functions can be used as a basis for selecting appropriate periods for continued service or for intervals of inspection and maintenance necessary to maintain reliability at an acceptable level.

Summarv of Results The methodology is illustrated using simple parametric representations of strength degradation and load process models.

Comprehensive data currently are being developed as part of the Structural Aging Program; here, we attempt to identify those parameters that have a particularly significant impact on time-dependent reliability so as to guide subsequent data acquisition.

Results to date indica % that the function describing the mean strength degradation and the mea occurrence rate of significant load events appear to be most important.

Factors that apparently are less important include the correlation between component strengths in a system and variability in the initial strength.

The hazard function, or conditional failure rate, is clearly nonlinear for the degradation mechanisms studied, and increases sharply after a prolonged service life for the degradation functions assumed in the analysis.

Thus the assumption of a linear failure rate would lead to an overly optimistic appraisal of reliability.

74

i Prediction of Aging Degradation of Cast Stainless Steel Cornponents in LWR Systerns*

Omesh K. Chopra Materials and Components Technology Division Argonne National 1.aboratory, Argonne, Illinois Cast stainless steels used in light water reactor (LWR) systems for primary pressure-boundary components such as valve bodies, pump casings, and primary coolant piping are susceptible to thennal embrittlement at reactor operating temperatures, i.e., 280-350*C.

Thermal aging of cast stainless steels at these temperatures increases hardness and tensile strength and decreases ductility, impact strength, and fracture toughness of the material.

Investigations at Argonne National Laboratory and elsewhere have shown that the, mal embrittlement of cast stainless steels can occur during the reactor design hfetime of 40 y.

Current assessment of thennal emorittlement of cast stainless steels involve simulation of end-of-design-life reactor conditions by accelerated aging at higher temperatures, viz.,

400 C. Estimates of mechanical-property degradation of cast stainless steel componcuts are based on an Arrhenius extrapolation of high-temperature data to reactor operating conditions.

A procedure and correlations have been developed for predicting the mechanical properties of cast stainless steel components during thermal aging in LWRs. The analysis focused on developing correlations for fracture properties in terms of material information in certified rnaterial test records and on ensuring that the estimated mechanical properties are adequately conservative for cast stainless steels defined by ASTM Specification A 351.

Fracture toughness of a specific cast stainless steel is estimated from the extent and kinetics of thermal embrittlement. The extent of embrittlement is characterized by the room-temperature Charpy-impact energy.

A correlation for the extent of embrittlement at

• saturation," 1.e., the minimum impact energy that can be achieved for the material afte,r long-term aging, is given in terms of chemical composition, Extent of thennal embrittlement as a function of time and temperature of reactor service is then estimated from the saturation impact energy and correlations describing the kinetics of embrittlement, whleh are given in terms of chemical composition and the aging behavior at 400*C. The fracture toughness J-R curve is obtained from the correlation between fracture toughness parameters and room-temperature impact energy. A common lower-bound J-R curve for cast stattdess steels of unknown chemical composition is also defined for a given material grade and ternperature. Increase in tensile flow stress is detennined from the kinetics of thermal embrittlement: the initial tensile properties must be known in order to determine the flow stress of the aged material.

Practure toughness J c and tearing modulus l

are obtained from the estifnated J-R curve and tensile flow stress, The correlations do not consider the effect of metallurgical differences that may arise from differences in production heat treatment or casting process and. therefore, may be conservative for some steels.

• Work supported by the olTice of Nue car Regul.itory Research. U.S. Nuclear Regulatory Commisaton FIN A22431.

Program Manager Dr. J. Muscara.

75 1

l l

Effect of Aging on the Predicted Maximum Load Carrying Capacity of Circumferentially Cracked Cast Stainless Steel Pipe Prabhat Krishnaswamy and Paul Scott BA'ITELLE 505 King Avenue Columbus,011 432012693 U.S.A.

ABSTRACT Cast stainless steel used in LWR primary system components such as valve bodies, pump castings, and piping is susceptible to thermal embrittlement at reactor operating temperatures,280-320 C (536-608 F). This process of thermal aging causes an increase in the hardness and ultimate tensile strength of the steci and at the same time a decrease in toughness. Work at Argonne National Laboratories (ANL) has shown that such thermal embrittlement due to changes in the microstructure can occur even during the reactor lifetime of 40 years. An Arrhenius type kinetic model has been used by Argonne to develop a correlation for predicting the mechanical properties (J-R curve and strength properties) of aged cast stainless stccl. The effect of this thermal

' degradation on the load-carrying capacity of circumferentially cracked piping is the subject of this work.

In this study, both lower bound and average values of the J R curve and the te nsile properties fc.r CF8M and CF8A cast stainless steel aged at 300 C for 20,40, and 60 years were used to predict the maximum load-carrying capacity of cracked pipe. Both through-wall-cracked (TWC) and surface-cracked (SC) pipe have been considered. The effect of aging, that is, reduced toughness and increased strength, for different pipe diameters and crack lengths has been investigated. Three analyses methods have been used to estimate the maximum load-carrying capacity of pipes: (1) a J-estimation scheme for TWC pipes developed by Paris that uses the J-R cuive as well as the yield strength and flow stress of the material,(2) a Plastic-Zone-Screening Criteria (DPZP) developed at Battelle which is applicable to both TWC and SC pipe which uses J and the flow stress, and i

(3) the R-6 Option 1 method developed by CEGB which uses the J-R curve, yield strength, and flow stress and is applicable for both TWC and SC pipe.

76

l Evaluation of Aging Degradation of Structural Cornponents' O, K. Chopra and W. J. Shack Materials and Components Technology Division Argonne National Laboratory, Argonne, Illinois Structural materials exposed to reactor environments undergo microstructural changes that can influence mechanical and corrosion properties and thus compromise the integrity of structure components. Current assessments of end-of-design-life mechanical properties of reactor components are mostly based on simulations by accelerated test conditions. Aged materials from reactor components offer an excellent opportunity to validate the laboratory studies. Mechanical properties of neutron shield tank (NST) material (A212 Grade B steel) from the Shippingport reactor, as well as of cast stainless steel (CF-8 steel) components from Shippingport and KRB reactors, have been charactertzed. The results are compared with estimates based on laboratory studles.

Increases in Charpy transition temperature (C'IT), yield stress, and hardness of the NST material in the low-temperature low-flux environment are consistent with the test and Army reactor data for irradiations at <232*C (<450 F). The shift in CIT is not as severe as that observed in High Flux lsotope Reactor (HFIR) surveillance samples; however, it shows very good agreement with the results for HFIR A212-D steel irradiated in the Oak Ridge Research Reactor (ORR). The results suggest that radiation damage in Shippingport NST and HFIR surveillance samples may be different because of the neutron spectra and/or Cu and Ni content of the two materials. High-flux low-temperature irradiation experiments are in progress on materials from the Shippingport NST and the IIFIR vessel to evaluate the possible effects of compositional and metallurgical differences between the two materials.

Mechanical-property data from cast stainless steel componens indicate relatively modest decreases in fracture toughness and Charpy-impact properties and an increase in tensile strength. The procedure and correlations based on laboratory-aging studies for estimating mechanical properties of cast stainless steel components in LWR systems, predict accurate or slightly conservative values for Charpy-impact energy, tensile flow stress, fracture toughness J-R curve, and J c of the Shippingport and KRU materials. The i

kinetics of thermal embrittlement and degree of embrittlement at saturation, i.e., the minimura impact energy that would be achieved after long-term aging, were established from materials that were aged further in the laboratory at temperatures between 320 and 400*C. The results were consistent with the estimates. The correlations successfully predict the mechanical properties of the Ringhals 2 reactor hot-- and crossover-leg elbows (CF-8M steel) after service of -15 y,

• Work supported by the omce vf Nuclear Regulatory Pesearth, U.S. Nuclear Regulatory Commission FIN A22562, i

Program Manager, Mr. E. Woolndge.

77 I

y U S NUCt(CR REGUL ATORY COMMIS$10N 1

1 LN It i NMyonM33b

~^

' ~ " " '

E'"$@

BIBLIOGRAPHIC DATA SHEET n,,,,,,,,m,

- e se,..m,;

NUREG/CP-0121 7.111 L L AND SUBlli L i Aging Research Information Conference - Abstracts of Papers

t

() All f4( POH i pub L thHE D vaNto tan March 1992 4 f IN OH GH ANI NUMM H

t. AVlHQH(h) b TYPiOfHIPUH1 Compiled by Ann Beranek Technical Conference ai nioo cuv t Hi o n,m..

v..,

4 Pl H f 0kMING QRG ant 2 A T ION + N AMt A ND ADDH t h Hf Nat. prwum D*sawm. ONe e' Nwon U S h ka **vsd. fury Comma *** *W "***8 **"'" " ' pa * 8"'

P'"'*

naec ces memns amoms 3 Division of Engineering Office of Nuclear Regulatory Research U. S. Nuclear Regulatory Commission Washington, DC 20555 0, SPONSOh tN3 ORG ANll A1lON ~ N AM L AND ADUH L S% O' kRC. fvVr %"* ** *h""* ' *' c*""* **t 9'"*

• N M C D" ** ** O'% * ** * *W*" USh****'"'****'"'>L*****"

~

and maame enknowstJ Same as above

10. SU?PLE ME NT ARY NOTE S

11. ABS T R ACT uoo.wm e,==>

This report contains abstracts of papers to be presented at the Aging Research Information Conference held at the Holiday Inn Crowne Plaza in Rockville, Maryland, on March 24-27, 1992.

This conference is held to disseminate research results in the area of nuclear power plant aging from programs sponsored by the Office of Nuclear Regulatory Research, U. S. Nuclear Regulatory Commission.

The conference will also provide an opportunity for engineers and scientists from around the world to exchange technical information and discuss future international cooperation.

The abstracts appear in the order in which they will be presented at the conference, and they are grouped by technical session.

The full papers and the agenda for the conference will be published as separato documents, u n c Y wa s osto e sc a w i o H s n, - --,.. ~< -,u m.-, ~ ~~ - ~.., ~ ~.-" >

" a anam"'"a = '

Unlimited Aging a $. cu. a. ua w...ca n*

Aging Research

,,~.e, Conference Unclassified m n

.n Unclassified Ib NUMttLHOFPAbtS 16 PHICt NQC 00mW 336 0 89)

~

1 THIS DOCUMENT WAS PRINTED USING RECYCLED PAPER t

[#

UNITED STATES g,c,y.oum ci as, mi NUCLEAR REGULATORY COMMISSION

• '^jan'"^*

WASHINGTON, D.C. 20555

{

,,,,,,, c y o

OFFICIAL DUSINESS 1

h.

PENALTY FOR PRIVATE USE. 6300

{ 1 b I+

JS PUBLII

~

gn.W-sunta cc r e

nhacu n

C>

f M f ge ze ng ce Q$

m Os m

7:

mC m2 i

l

. _ - - _ _ - _ _ _. _ _ - - _ _ _ _ _ _ _ _ _ _ - _ _ _ - _ - _ _ - - - - _