ML20116M888

NRR Technical Newsletter.Volume 2,Number 1

ML20116M888
Person / Time
Issue date:	04/30/1990
From:	Office of Nuclear Reactor Regulation
To:
References
NUREG-BR-0125, NUREG-BR-0125-V02-N1, NUREG-BR-125, NUREG-BR-125-V2-N1, NUDOCS 9608210145
Download: ML20116M888 (8)

Text

_ _ _

i y

a 9

1

(

OFFICE OF NUCLE AR RE ACTOR -REGULATION

=

j U.S. NUCLEAR REGULATORY COMMISSION i

a e,

,a l,u m

a m

1==

i i

TECHNICAL nuase/sa.0,2s NEWSLETTER MMM

)

Well Done, NRR Staff l

by James H. Sniezek i

l In the April 1987 reorganization,we faced a great challenge in control room design review (DCRDR) reflect the actions of a con-i making the transition to a vibrant organization with the main fident safety staff.

l focus of operational safety. Together we had to learn how to j

combine the best features of the licensing, operational safety Your reaction to the daily events that occur in the plants, ranging i

review, and inspection staffs into a single coherent operation.

from routine licensee event reports (LERs) to the need to staff the Operations Center, reflects a commitment to the important role You have succeeded in a remarkable fashion. You have taken you have in ensuring the safetyof the plants. You are consistently control of the backlog of licensing actions and have put ensuring that the Agency reacts properly in communicating safety j

measures in place to prevent the growth of a backlog in the information to or imposing requirements on licensees.

4 future. You have a good traceable knowledge of the status of Three Mile Island actions and of Unresolved Safety Issues and At the same time, you are demonstrating a sensitivity to the basic have comprehensive actions under way to fully understand the tenet of our mission: "The licensee is responsible for the safe status of generic safetyissues. You have made good progress operation of the facility. Ourjob is to ensure the licensee fulfills his on new programs such as plant life extension and evolutionary responsibility." You have demonstrated a willingness to allow and reactor reviews.

encourage the licensee in his mission through the proper use of the provisions of the regulations in 10 CFR 50.59; a willingness to allow The knowledge gained by NRR and the Agency of the opera-the industry to solve its problems in areas such as training, in-tional safety of plants is at an all-time high because of your service testing, as low as reasonably achievable (ALARA), fitness visits to the plants and firsthand observations of the effect of for duty, and maintenance; and a willingness to cause plant im-your reviews and decisions. This knowledge, combined with provements in consonance with the backfit rule in areas such as the your participation in and leadership of the inspection program adequacy of the auxiliary feedwater system.

,and the systematic assessment of limnsee performanc (SAlf) program, and your preparation of NRC managers for the You have taken the initiative to resolve many nagging issues that semiannual discussion of problem facilities, has resulted in an have persisted for quite some time. Accident management, leak-increased awareness of safety performance.

before break, severe accident resolution, and containment per-formance are areas in which your dedication to solving difficult You have reached new heights in bringing to the NRC a tasks is apparent.

renewed awareness of the role of human factors and probabil-istic risk assessment (PRA) in the achievement of operational C*"d"a en Iw 2 safety. You are reacting from the perspective of safety and not i

merely from the perspective of compliance. Your new ap-What' sin thisissue?

a i

proaches in the resolution of operator licensing issues, multi-O)I~'

plant actions, and longstanding open generic items such as the l

ggg pagg p safety parameter display system (SPDS) and the detailed 9608210145 900430 1

PDR NUREC BR-0125 R PDR

p y

stcffs, you have demonstrated that by working together in cn atmosphere of mutual respect and coope. ration we can make (;reat strides in improving the quality and timeliness of the office prod-J IN THIS ISSUE

+

ucts. Your prmespal correspondence record is superior, and your management and tracking systems are constantly improving.

Well Done NRR Staff byJames H.Sniezek ;

1 Although you have accomplished a great deal, do not become complacent. Instead, carefully evaluate your own work, search out Use of Performance Indicators for Measuring new areas for improvement, and take each criticism as a challenge i

Engineering Ouality to do even better. You can be proud ofyour accomplishments and by Gerald E. Gears and Robert C. Pierson, ADSP 2

look forward to even greater success in your individual and collec-tive endeavors. As the Deputy Director,it has been a true honor Inspecting a Raw Cooling Water System at for me to work closely with you over the past 3 years and I thank you Sequoyah for MicrobiologicallyInfluenced for all that I have learned from you.

Corrosion Damage by John York, Region II.-...

4 You are truly professionals.

Cable Damage at Watts Bar by E. C. Marinos, ADSP 5

Use of Performance Indicators Regulatory Significance of the Diablo Canyon FOr Measuring Engineering Long-Term Seismic Program (LTSP)

Quality by Harry Rood, DRSP 6

Brittle Fracture of Aluminum Bronze Casting by Gerald E. Gears and Robert C. Pierson, ADSP by Herbert Kaplan, Region I......

7 The NRC in its role as the principal safety regulator for nuclear Potential Loss of Required Shutdown Marg. Durm.g power reactors has traditionally preferred that licensees propose m

PWR Refuchng solutions to both program and licensing problems. Generally, the by Laurence I. Kopp, DST.~..

8 NRC chooses not to dictate solutions to problems that evolve from the operation and control of power reactors. However, over the NEWSLETTER CONTACT:

past few years, the NRC has chosen to become more active in Valeria Wilson, NRR 492-1208 identifying plant problems through the use oflarge team inspec-j tions. These inspections provided some interesting results.

The following case depicts a licensee's approach in resolving sig-nificant hardware and n n-hardware weaknesses which resulted in Well Done NRR 8

an extensive shutdown period and mcreased regulatory attention.

l In this case, the plant operated by the licensee was placed in a Comiinued from Page I shutdown mode and was identified as requiring heightened regu-latory attention. Because of concerns raised by the NRC staff and Your technical competency and professionalism are recog-by its own management, the licensee began an extensive program nized by all segments of the Agency. You are in demand to to correct a wide variety of deficiencies in the areas of hardware, 1

serve on and lead inspection teams, and serve in rotational operations, and management. This licensee identified two major assignments on the staffs of the Office of the Executive Direc-causes of its deficiencies: (1) the lack of sufficient numbers of tor, the Commission, the regions and the Office of the Secre-experienced nuclear managers, and (2) the absence of an effective tary. When a tough issue arises, the general tendency is to organizational structure to ensure safe construction, operation, "give it to NRR, they can solve it."

and management ofits nuclear pbnts. To alleviate its deficiencies, the licensee undertook many self-initiated ccrrective action pro-You have sought out and grasped new ways of doing business grams to resolve both hardware and organizational problems.

in an effective and efficient manner. You completely revised the inspection program and inspection techniques, applied A recurring deficiency in both the development and implementa-PRA insights to our many daily activitics, solved the many tion of previous corrective action programs had been the inability contentious issues of emergency preparedness, streamlined of thelicensee to articulate and measure the problem that a change our resource management systems, and identified training was intended to correct or improve, the expected result once the opportunities that will increase the continued professionalism change wasimplemented, and evidence that the change did,in fact, of the entire NRR staff. In the tecimical and administrative result in the desired correction or improvement. The nuclear 2

L

l

~

industr is not unique inthis Ifick of ability to express a standard percentage of safety evaluations accepted by the o

by which to measure and monitor program effectiveness. Suc-Plant Operations Review Committee (PORC).

cessful organizations a're able to develop and define goals and objectives, provide the necessary resources to attain these goals The process of choosing and monitoring indicators will, of neces-and objectives, and provide a fair nechanism for assessing both sity, be an evolutionary process. The description of the second the process and the results. This framework is essential for an indicator illustrates the organizational dynamics involved in creat-effective organization.

ing and modifying an indicator. The number of field changes needed for each design change package is both simple to quantify To upgrade the quality control of its nuclear engineering group, and is intrinsically objective. When a design change package is the licensee established a separate engineering assurance (EA) delivered to the field, the personnel responsible for implementa-group. This group proved very effective in identifying major tion must be able to perform the change as specified. If, for some weaknesses in the nuclear engineering group and establishing a reason, a change must be modified or is rejected, each noted basis for quality control audits. The EA group provided broad, difference is counted as a field change. This indicator is important corporate-based verification for both line engineering responsi-because it provides data on the ability of engineers to correctly bilities and typical functions for the oversight of quality assur-translate required changes into a package that can be imple-ance. From the ongoing assessment by the NRC of the licensee's mented. By necessity, the engineer must have a good understand-improvement programs and managerial changes, the NRC ing of the requirements and also must account for plant conditions consistently rated Ihe EA group as an effective organization for and equipment interface. Failure to do so results in one or more quality oversight. However,once severalimportant milestones licld changes and is an indication of the engineering staff failure to were achieved, the licensee chose to integrate the EA quality adequately perform their assigned tasks. Therefore, this indicator function into the engineering line organization and chose to provides an excellent example of a direct measurement of quality integrate the oversight function back into the nuclear quality as-performance.

surance group, thereby disbanding the EA group. The NRC was requested to review and approve this change as part of the licen-When implementation is assessed, the efforts to develop quantita-see's continuing effort to resolve identified managerial and tive measures are proving to be an effective tool for corporate man-organizational weaknesses. Although the NRC understood the agement to monitor the results of their organizational decisions. In importance of transferring the responsibility for quality to the addition, this approach provided an effective means for the NRC i

line organization, it was concerned that the excellence devel-to encourage licensees to develop, validate, and use quantifiable j

oped in the EA group would be lost or diluted by Ihis change. indicators to assess the results of their own decisions, when appro-priate. Also, this approach provides the NRC with measurable Without the necessary framework in place, including goals and data oflicensees' performance. Encouraging the development of objectives, provisions for resources to attain them, and measure-simple and direct measurements of what is important provides an ment tools for assessing the results, the licensee could not extremely valuable tool to the NRC and the licensee for the overall determine which implemented changes were beneficial. With-assessment of effectiveness of changes directed by management.

out any methods to verify this necessary cause and effect rela-These assessments should lead to improved operations, regulatory tionship,it became apparent that the licensee would continue to com pliance, and safety. In summary, recent experience of the NRC be unsuccessful in its search for an acceptable solution to its and the licensees indicates that some of the guiding tools so needs to effectively use staff and resources. Therefore, the NRC successfully developed in non-nuclear industries to promote high encouraged the licensee to develop performance indicators. quality performance can be applied in the nuclear industry. This These indicators would become part of the evaluation process circumstance appears to be especially true in the area of critical that had been missing from previous changes; a process to self-assessment. Either as a result ofincreased regulatory attention measure performance in a simple but quantitative fashion. This or of their own critical self-assessment, many utilities are making performance was the success or failure of the changes as com-strong commitments to increase the quality of their operations and pared to their intended goals, objectives and consequences.

are striving to increase overall safety. However, as shown in this case study, quality commitments must be translated into goals and Following discussions with the NRC, the licensee began the objectives and a guiding system must be established to include the development of several quantitative measures to track the prog-measurement of progress toward this end. As illustrated previ-ress of the integration of EA into quality assurance (OA) and, ously, the development of simple, quantifiable measurement tools more importantly, the progress of its engineering group in is often the missing key. More importantly, if these tools are to providing high-quality products to the construction group. The work, they should be developed at the beginning of a program, they licensee chose three indicators:

should be visible to all, including line management and senior-level management, and they should be simple and relatively few in percentage of satisfactory nuclear engineering number. Ultimately, the goal should be that measurement of o

deliverables; quality will be done more "by" the participants (that is, the team, the department, etc.) rather than "to" the participants.

number of field changes for each design change o

package; and 3

w

F l

Incpacting a R0w Cosling s"fety-re ated system.

~

Water System at Sequoyah DEVELOPMENT OF AN INSPECTION PROGRAM For M.icrobiologically TVA decided to perform a semiannual, visualinspection of each influenced Corrosion Damage or the approximately 400 stainless steel welds in the ERCW system. If a leak should be detected, the weld would be radio-by John York, Region ll graphed within 7 days. If the length of the corrosion damage exceeded a certain liaear value (837 inches for a 6-inch diameter INTRODUCTION pipe), then an evaluation fracture toughness would be performed to determine if an immediate repair would be r.eeded or whether This case history was collected during several inspections of the the particular loop of the ERCW system could continue to oper-l efforts of the Tennessee Valley Authority (TVA) to assess the ate. If a leak was detected in a weld, then the weld would be condition of the essential raw cooling water (ERCW) system at visually inspected each month and radiographically evaluated l

TVA's Sequoyah Nuclear Plant on the Tennessee River. The every 3 months until the weld was replaced. If a weld was found l

ERCW system consists of two independent and redundant loops to be leaking, it would be replaced at the next refueling outage.

within each unit. This system used the Tennessee River for both suction and discharge. in 1980, the utility replaced part of this The ERCW system was inspected one loop or train at a time. Each system's smaller-diameter carbon steel piping with a 316 stain. loop has a cooler pipe (suction side) that is insulated with rubber less steel piping. The pipes made of carbon steel at several of the foam and a warmer pipe (discharge side). Although the welds can utility's nuclear plants were experiencing flow restrictions be. be easily detected on the uninsulated warmer pipe, welds on the l

cause of the buildup of nodules of microbiologically influenced cooler pipe have special coverings and are located for inspection corrosion (MIC). The utility made a decision that this condition by the use of weld maps. Actually, each weld on each of the could be eliminated by use cf stainless steel piping.

stainless steel pipes became part of a formal preventive mainte-nance program so that a record could be kept of the inspections DISCOVERY OF MIC A'ITACK and results. Some of the welds were physically difficult to inspect because of the closeness of other pipes and equipment, closeness During an inspection of the ERCW system for leaks because of of electrical cable trays, and the height above the floor requiring l

corrosion conditions encountered at one of the other TVA scaffolding to be erected in order to inspect the piping.

nuclear plants, operators discovered three leaking welds in a part i

of the ERCW system. The stainless steel piping for this system D uring each visualinspection the TVA inspectors would walk the had approximately 400 butt welds. Once these leaking welds length of all of the stainless steel piping and would examine any i

l were discovered, a decision was made to examine some of the discolored areas around the welds or any visible signs of leaking other welds. Initially, to create a surface appropriate for an water. Sometimes the rust-colored stains around a weld were the ultrasonic inspection, engineering and inspection personnel decided result of a leaking flange or valve located above the weld. These to fletten the weld crown on 27 welds by grinding. Of these 27 areas were cleaned, and any small dark brown areas were exam-weldr.,13 developed minor leakage after grinding indicating that ined further with a metal probe. Experience developed at TVA l

MIC corrosion had penetrated most of the weld metal.

showed that dislodging the dark brown corrosion product from the l

MIC penetration would usually make the area leak again.

l Personnel removed a horizontal section of 6-inch diameter piping welded to a 90-degree elbow, from the affected piping for TVA inspectors found no steady streams of water coming from the evaluation. A number of MIC nodules were located at various MIC areas. The worst case discovered durit iese inspections po4tions around the circumferential welds. Removal of these involved a leak rate of approximately 60 d

.per minute. Two MEC nodules from the inside diameter of the piping revealed days later, an examination of this weld revt.ed a leak rate of10 to concentric rings of discoloration with small light areas near the 15 drops per minute. It appeared from these inspections that a center of the area. These small light areas were determined to leak would initially develop from a MIC penetration. Later, the be the openings of corrosion pits that provided entry to larger leak rate would decrease and eventually stop because the buildup l

sutiaurface corrosion areas.

of corrosion product in the penetrations. This eventual stoppage appared to happen in spite of fluid pressures of 90-100 psig and TVA personnel found that a radiographic inspection produced fluid temperatures ranging from 50 to 80 degrees Fahrenheit on more information on the extent of MIC damage than ultrasonic the suction, and 95 to 100 on the discharge side. Also,it was noted testir.g produced. Therefore, they adopted this method for during these inspections that the leaks were not only located on furthcr evaluation. The utility personnel radiographed 67 welds horizontal sections of the piping but on vertical runs of piping as and found 61 had MIC damage. Not all of the 400 welds were well. Leaks were found at various positions around the circumfer-radioguphed and evaluated, and therefore, an inspection pro. ence of the piping. No significant difference was noted between gram had to be developed to monitor further attacks on this the number ofleaks in the cooler intake piping and the number of 4

f

leaGin the warmer discharge P P ng.

gg Nmp g gg g ii During one of these inspections, an examination of a weld in the by E. C. Marinos, ADSP vertical position of a 3-inch diameter pipe in the ERdV system revealed a very small drop of water. When the water was wiped Cable installation practices at the Watts Bar Nuclear Plant came off,it did not reappear until l hour later, indicating a very small into question by the NRR as a result of the concerns of TVA l

amount of leakage. Clearing the area with a metal pick revealed employees regarding inadequate procedures for cable installation i

a MIC penetration through the wall and caused the area to start and violations of existing procedures. In June of 1989, TVA, l

leaking water again. Radiographic examination of this weld through an employee concerns program, removed the cables from l

showed extensive subsurface MIC damage. This damage re. a conduit run in the reactor protection system of Unit 2 to inspect quired an analysis of fracture toughness before the weld was for damage. This conduit was selected in response to an employee qualified for use without repair. This analysis revealed that the concern that a welding arc that struck the conduit during construc-extent of MIC damage cannot be directly correlated to the tion may have damaged cables in the conde:t. When the cables amount ofleakage.

were removed, significant damage was found in the insulation of some cables. However, this damage was not attributed to heat TVA decided to use radiography to monitor the MIC attack on generated by the alleged welding arc. The damage was principally three old welds and three new welds in this system. The goal was attributed to the pulling stresses exerted during the initial installa-l to try to establish a MIC attack rate to substantiate or to change lion of the cables.

the existing frequency of inspection. The old butt welds con-sisted of 316 stainless steel piping welded with 308 stainless steel At Watts Bar and the other nuclear facilities at TVA, a common l

filler metal. These welds had been in service for approximately method was used to install cables in conduits. To fill a conduit, pull 9 years. Two of three of the welds had been attached by MIC at cords were used to pull additional cables through the conduit over i

I the start of the monitoring period. The newbutt welds had been the top of existing ones in the conduit (pull-bys). Potentially this welded with 316 stainless steel filler metal and had been in practice can cause damage to the existing cables from the sawing l

service for 6 months. Using radiography,it was determined that action generated by the pull cords and by the cables themselves as two of three of these welds had been attached by MIC at the they are pulled over the existing cables. Usually damage can be 4

l beginning of the monitoring period. Approximately 1 year after avoided by using adequate amounts of lubricants, by controlling these results were established, all six welds were found to have Pulling tension, by choosing appropriate pull cords, by controlling 4

been attached by MIC to some degree. The results of this the distance between pull points, and by minimizing the number specific monitoring did not provide sufficient information to and angle of bends allowed in the conduit run. Presently, the l

determine a rate of attack.

industry standards provide no specific guidance for performing l

multiple pulls of cables in conduits. The concerns raised byTVA l

Part of the inspection process for the nuclear plant was to ensure employees and the NRC staff have heightened industry interest in l

that the Class 1E electrical equipment and instrumentation this subject. As a result, working groups have been created to necessary to safely shut down the plant was protected from dcVel0P SPecifie guidelines for dealing with multiple cable pulls in

)

l leakage and potential water damage. Electrical prints were used conduits in the future, i

to determine the location of electrical equipment in the area of l

ERCW stainless steel welds. The utility decided to place pro. To assess the adequacy of cable installation at its nuclear facilities l

tective boots around the welds in the vicinity of this electrical TVA instituted programs for corrective action. At Watts Bar, equipment. The boots were composed of plastic sheet material overlays of the damaged cables on plant isometric diagrams of held on each end by a stainless steel band and sealed in the joint conduit runs have indicated that cables appear to have been areas by a silicone-rubber compound. The boot also had a drain damaged at locations of the conduit runs where pull tensions and of clear plastic tubing so that any leaking welds could be side wall bearing pressures (SWBPs) have exceeded certain safe detected.

threshold values. The TVA program for corrective action calls for replacement of cables that have exceeded the threshold values of Not all plants are susceptible to attack by MIC. However,in pull-tension cad SWBP. Pull tension and SWBP values are those plants that are, a continuing, organized program must be calculated as a function of the physical parameters of the cables and maintained to ensure the structuralintegrity of the system. To the conduit configuration. TVA's cable installation procedures accomplish this, the approach taken may vary between nuclear (G-38) included conservative values of SWBP that the cable plants. The approach used will depend upon the extent of the installation crews may not have followed at the time of major problem, the type of damage, the rate of attack, and the acces. construction at Watts Bar.

sibility of the affected components. In this case, acute damage 1

was confined to butt welds in the stainless steel portion of the For the Browns Ferry Nuclear Power Station, TVA is applying an ERCW system. These welds could be readily inspected during approach similar to that of Watts Bar for identifying conduits with any mode of operation of the nuclear power plant operation.

cables that received the highest amounts of pull tension and SWBP during installation. Ilowever, to assess the adequacy of the instal-l lation of cables at Browns Ferry, the proposed program for correc-

tive action requires that cables be tested using a direct current fornia. For this analysis, the licensee ussd tide gauge re,wrds from *

(de),high potential testing method. This testing method applies San Francisco and San Diego and a computer simulation of the to cables in a group of conduits where calculations revealed the propagation of the tsunami waves produced by the earthquake.

l highest values of pull tension and SWBP. Before testing, the conduits will be flooded with water so that a firm electrical ScurmcSource 0:aractenzation ground can be maintained between the conductors and the external surface of the insulation. Failure of the cable to For each fault in the plant vicinity, PG&E developed a maximum withstand the high potential test willindicate that the insulation magnit ide and a representation of magnitude as a function of may have lost the required dielectric strength because of physical probability. This effort to characterize the seismic sources at damage caused by the pulling stresses exerted during installa-Diablo Canyon involved new analytical approaches using the most tion. This test will provide statistical data for determining the extensive data base ever used for similar studies conducted for a adequacy cf the procedures for cable installation used through-nuclear power plant.

out the facility at Browns Ferry.

l Deterrmnatwn ofGround Mo %

l Rcgulatory Significance of the The licensee conducted studi.:s of ground motion to determine the l

Diablo Canyon Long-Term spectral accelerations in the froe field at the plant site that would result from earthquakes on faults in the plant vicinity. PG&E Se.ismic Program (LTSP) conductea regression analyses of empirical ground motion data to l

determine an equation describing the attenuation of spectral by Harry Rood, DRSP acceleration with distance from the earthquake. Similar studies were also conducted by the U. S. Geological Survey, acting as a The Long-Term Seismic Program (LTSP) for the Diablo Can-consultant to the NRC. These empirical regression analyses used yon Nuclear Power Plant is a several-year, multi-million dollar the most comprenensive data base of earthquake recordings ever l

cffort bythelicensee,PacificGas and Electric Company (PG&E), used for nuclear power plant evaluations. In addition, PG&E l

to reevaluate the seismic design basis for Diablo Canyon. The performed numerical earthquake analyses using a finite element original recommendation for the reevaluation came from the model of the earthquake fault and the surrounding area. This l

NRC's Advisory Committee on Reactor Safeguards, which stated modelsimulated the earthquake fault rupture and the propagation l

in its July 14,1978, letter approving operation of Diablo Canyon: of seismic ground motion to the site. These models were then "The Committee recommends that the seismic design of Diablo applied to the Diablo Canyon site using the site-specific informa-Canyon be reevaluated in about 10 years taking imo account tion developed by the geologic and geophysical parts of the LTSP.

j applicable new information." This recommeadation was made The ground motion studies also provided a range of data on ground a firm regulatory requirement by its incorporation as a license motion for use in the probabilistic risk assessment (PRA) of the i

condition into the full-power operating license for Diablo Can-scismic hazard at Diablo Canyon.

yon Unit 1 that was issued on November 2,1984. To comply with the license condition, PG&E initiated the LTSP in 1985 and Soit-SimctureIntemctwn (SSI) Analynr l

issued its final report on the LTSP on July 31,1988. The NRC staffis reviewing the LTSP final report and expects to con.plete The LTSP 3-dimensional SSI analysis used site-specific geologic the review in 1990.

information and included a consideration of the effects of the incoherence of ground motions and the uplift of the containment l

The LTSP is broad in scope, encompassing all technical disci-base mat.

plines necessary to characterize the adequacy of the seismic design of Diablo Canyon. The LTSP and the review ofit that the Probabdistic Risk Assessinent NRC staffis conducting have set new standards for geoscientific investigations and analyses for nuclear plants. Major efforts To put the seismic risk in perspective, the LTSP cffort included a were undertaken in the following areas:

complete levell PRA for Diablo Canyon (DCPRA). The DCPRA considered accident initiators other than scismic initiators,includ-Geolope and GeophysscalInnstipatwns ing other external and internal events. The seismic part of the DCPRA was the most detailed and extensive seismic PRA ever PG&E conducted geologic, seismologic, geophysical, and tec-undertaken and represented a major step in advancing the state of tonicinvestigationstoestablishthelocation, seismicity, type,and the art of regulatory analysis. Important features of the seismic capability of faults in the general vicinity of the plant. These part of the DCPRA included the development of detailed seismic investigaGons developed basic data,information, and interpre-hazard curves (probabilistic representations of the carthquake tations that are the most comprehensive and detailed ever devel ground motions at the site). These hazard curves were developed oped for nuclear plant licensing. Many of these investigations from (1) a field investigation of geology and geophysics; (2) a 4

used new evaluation techniques, such as an analysis to determine detailed structural analysis, including a non-linear analysis and the location of the 1927 offshore earthquake near Lompoc, Cali-soil-structure interaction; (3) a detailed fragility analysis of struc-6

tures ahd components'(a pfobabilistic representation of the ca-sure to a corrosive me'dium (sea water) over several years, the apab!!ity to withstand peismic motion); and (4) a detailed evalu-material dealloys as the aluminum is lost. This process results in 1

ation of relay chatter effects. While the DCPRA was used to the formation of a low-strength, copper-rich network. This eutee-dete mine the seismic margins at Diablo Canyon probabilisti-told condition was found in a relatively large section in the vicinity cally,information developed for the DCPRA was also used in of the end flange, supporting the conclusion that the weakness was the deterministic assessment of the seismic margin at Diablo introduced by casting and not by some other process (e.g. welding).

Canyon.

Metallography indicated that the fracture area consisted of a substantially interconnected eutectoid region that was significantly In summary, the LTSP resulted in a major advance in methods more susceptible to dealloying. Metallography also indicated that of evaluating the seismic design basis of nuclear power plants acceptable material contained the expected microstructure with and will have a significant effect on future geoscientific studies small amounts 5 percent of eutectoid that was not interconnected for nuclear power plants.

and was much less susceptible to degradation. Slow cooling of the casting in the flange is believed to cause the eutectoid condition.

Brittle Fracture of su nd fad r c ntributing i the fracture was the buildup 1

from the weld that had been performed in the upper flange region Aluminum Bronze Casting of this column to restore a spider bearing support surface that had been worn because of operation. Subsequently, a crack developed by Herbert Kaplan, Region I in the weld buildup. This crack was a pre-existing throughwall flaw with a maximum length of approximately 5 inches. Examination determined that this crack had existed for some time before the On January li,1990, during maintenance activities on a salt service water (SSW) pump at the Pilgrim Nuclear Power Station, application of the impact load caused by dropping the pump. The workers dropped one end of the approximately 43-foot-long Primary origin of the fracture was at the weld buildup. From there detached pump column assembly 6 to 12 inches onto a concrete the fracture propagated m a brittle fashiors through the low-floor. Upon impact, the column fractured 360 degrees in a strength, copper-rich network under the stress mduced by the brittle mannet. Sabsequent testing revealed severely degraded unp ct on the concrete.

1 material properties in localized areas of the column section. The columns are made up of seven sections that are bolted together The failed pump (A).is one of five s.imilar pumps employed at at flanged joints. The column sections were cast from an Pilgrim. The others (B, C, D and E) were operable at the time oi aluminum bronze alloy (ASTM B148, Alloy C95200) and were the failure. Following the failure, extensive field m, spections and i

fabricated by Ampco Metals. The pumps are Section - VIT 16 hardness tests were performed on the accessible areas of all DHLC single stage pumps manufactured by Gould Pumps, columns,,ncluding the new columns, the warehouse. No cracks i

m Vertical Products Division.

r areas of reduced strength (based on hardness) were uncovered.

Fracture tests and analyses ofIhe materialwith degraded mechani-Cal roperties were also performed to determine the material's P

A metallographic examination of the fractured area showed that the failure originated from a pre-existing defect near an area that ability to withstand loadmg with pre-existmg flaws. The tests and had been welded to refurbish a worn surface where sp* der analyses revealed that the degraded material had very poor resis-tance to fracture. Because of the wherent brittleness of time bearing supports contact the inside of the column. The welded area and the adjacent base metal showed evidence of porosity. material, the following structural changes were mitiated unmedi-Portions of the fractured surface also showed evidence of worm ately to provide an additional margin of safety. Lateral stabilizers hole, corrosion-likedefects. Tensile testingof materialremoved were added to two pumps and tie rods were added between the from the area adjacent to the fracture end showed low ductility, Hariges of the top two column sections of each pump. (Only two 1/2-percent clongation, and a low tensile strength of 23,000 psi, Pumps are required to satisfy the design basis accident). In the w rst case, the maximum stress levels were determmed to be below compared to minimum specified values of 20 percent and 65,0C0 psi, respectively. Testing of areas far removed from the fracture the American Society of Mecharu, cal Engmeers (ASME)Section V Code allowable for brittle material such as cast, iron.

revealed acceptable mechanical properties. Testing also showed a coirelation between hardness, ductility, and tensile strength. Fm the long term, all columns will be replaced with th The areas with degraded properties had a Brinell hardness of 75 to 90 compared to values of 105 to 135 in areas exhibiting accept-corrosion-resistant ahoy mckel alummum bronze alloy C95800.

The new columns w,ll be fabncated from centrifugally cast sec-i able properties.

tions. Installation of the new columns is scheduled to begin on June 30,1990. For the short term, each of the other pumps (B, C, The metallurgical findings indicated that two factors contributed to the fracture. Fix s the initial casting process probably pro-D ar.d F) will be removed, mspected on a programmed bas,s, and i

duced a susceptible in crial since metallurgical analysis identi-returned to service. On February 13,1990, pump A was returned fied a eutectoid conditsm in the fractured area which indicated t servicewith a refurbished column section. Investigation byNRR that the required single phase of the material was not achieved. indicated tha. the Pilgrim station was the only plant supplied with In thisinitial eutectoid condition,an aluminum-rich phascof the Pump columns fabricated from this material.

material forms at the grain boundaries. With continued expo-7 j

~

t

.4 F

Pst3ntial Loco cf R7quirGd 1.

Assure that any intermediate fuelpssembly configuration Shutdown Margin During.

(including control rods) intended to be used during refueling is identified and evaluated to maintain sufficient refueling boron PWR Refueh,ng concentration to result in a minimum shutdown margin of approxi-by Laurence 1. Kopp, DST 2.

Assure that fuelloading procedures allow only those inter-As a result of longer fuel operating cycles, utilities have been mediate fuel assembly configurations that do not violate the allow-increasing the enrichment of reload fuel. Some of these fresh able shutdown margin and that these procedures are strictly ad-reload assemblies may be highly reactive under certain refueling hered to.

conditions. For most core refuelings, operators move fuel di-rectlyinto the final core positions. However,when the spent fuel 3.

Assure that the staff responsible for refueling operations is poolbecomestoocrowdedtoaccommodateafullcoreinventory trained in the procedures recommended in Item 2 above and of fuel, fuel may be shuffled within the core before a final understand the potential consequences of violating these proce-configurationis achiewd. Duringthisshuffling,someirradiated dures. This training should include the fundamental aspects of fuel from previous cycles is discharged and exchanged for fresh criticality control with higher enriched fuel assemblies.

fuel. The concentration of boron at refueling is analyzed for pressurized-water reactors (PWRs) to confirm that the concen-Some of the respondents to the bulletin indicated that during tr: tion is sufficient to maintain the required shutdown margin refueling at various plants, some of the intermediate fuel assembly (subcriticality) for the final core configuration. However, these configurations may not have been analyzed previously. Even for smalyses may not be sufficient to assure that the shutdown plants that do not use in-core shuffles during refueling, there were margin will be maintained for all intermediate fuel assembly situationswhere theproceduredeviatedfrom theonloadsequence pc itions. In addition, because a significant amount of reactivity pattern and where assemblies were placed in temporary interme-c a be added to subcritical configurations by the addition of a diate positions. For example, Duke Power engineers identified the single highly reactive assembly, the reactor could inadvertently possibility of unanalyzed intermediate fuel assembly configura-reich criticality if a number of such assemblies are grouped tions. As a result, they performed criticality analyses on different together. With this highlyreactive fuel, indication of sub-critical clusters of 4.0 weight percent U-235 fresh fuel assemblics in 2000 multiplication by inverse count rate may not provide adequate ppm concentrations of borated water. These assemblies did not warning of an approach to criticality. An inadvertent criticality contain any control rods or burnable absorber assemblics. For the could result in fuel failures, system damage, and high radiologi-interior regions of the core, the analyses determined that any cal doses to onsite workers, cluster of more than four fresh fuel assemblies could violate the allowable 5 percent shutdown margin and should be avoided. This Baltimore Gas and Electric Company (BG&E) submitted a 10 data confirmed the results reported by BG&E that led to their 10 CFR Part 21 report to the NRC on March 15,1989, regarding CFR Part 21 notification.

the potential loss of shutdown margin during refueling opera-tions at its Calvert Cliffs Nuclear Power Plant. In addition, Com-The NRC decided to send Bulletin 89-03 only to PWR licensees, bustion Engineering, Incorporated (CE), the consultant for and decided that licensees of BWRs should not have to respond to nuclear fuel design for Calvert Cliffs, recognized this potential the bulletin. The primary reason was that, because of the imbed-problem and issued an information bulletin to all utilities with ded burnable absorbers in the BWR fuel matrix, the maxirnum re-CE-designed plants. In circumstances in which explicit analyses activity of BWR fuel does not appear to increase with higher en-l are not available for each intermediate fuel assembly position, richments as much as the maximum reactivity of highly enriched CE recommended positioning fuel only in intermediate core PWR fuel increases. PWR fuel may or may not contain burnable locations that will contain fuel of equal or greater reactivity in the absorbers. In addition, the NRC determined that the BWR final core configuration. BG&E revised the refueling proce-licensees had been given adequate information to consider these J

dures at Calvert Cliffs to allow fuel to be positioned only in problems through General Electric safety information letters and intermediate core locations that will contain fuel of equal or previous information notices on the same subject.

reater reactivity in the final core configuration.

The Committee to Review Generic Requirements (CRGR) rec-l In response to the BG& E notification, the NRC issued Informa-ommended the exploration oflonger-term solutions to the prob-I tion Notice No. 89-51, " Potential less of Required Shutdown iem described in Bulletin No. 89-03, including stronger measures MarginDuringRefuelingOperations,"datedMay31,1989,toa11 than administrative controls. As a result of these recommenda-holders ofoperating licenses or construction permits. This infor-tiens, NRR has requested technical assistance from one of its j

mition notice required no specific action or written response. contractors to perform analyses. Differing refueling configurations Subsequently, the NRC issued Bulletin No. 89-03, dated Novem-will be analyzed to understand the potential for losing shutdown ber 21,1989, which requested that all PWR licensees take margin and for inadvertent criticality. This information will then l

specific actions. The following actions were requested:

be used to help assess the need for any additional requirements.

I 1

i

- -. -