IR 05000317/1990017

Insp Repts 50-317/90-17 & 50-318/90-15 on 900710-13.No Violations Noted.Major Areas Inspected:Review of Several self-identified Welding Deficiencies Involving Unit 2 Pressurizer Sleeve Replacement Program

ML20059E345
Person / Time
Site:	Calvert Cliffs
Issue date:	08/16/1990
From:	Kaplan H, Lohmeier A, Terao D NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I)
To:
Shared Package
ML20059E338	List: IR 05000317/1990017 ML20059E337
References
50-317-90-17, 50-318-90-15, NUDOCS 9009100149
Download: ML20059E345 (10)

Text

, m

+ . ,. .

.. .

..- .

.

.

UNITED STATES NUCLEAR REGULATORY COMMISSION ,

REGION I

-Report No.: 50-317/90-17 50-318/90-15 Docket No.: 50-317 50-318 License No.: DPR +'i3 .

)

liF69 '

Licensee: Baltimore Gas and Electric Company MD Rts 2 & 4, P.O. Box 1534 '

Lusby Maryland 20657-Facility Name: Calvert Cliffs Nuclear Power Plant U Inspection At: Lusby, Maryland pP l

Inspection Dates: July 10 to July 13, 1990 '

.

i Inspectors: / W M g d .pn Herbert Kaplan, S .c , tor Engineer, Materials date and Pr sses See a, EB, DRS

_

Alfred f6hmeier,' Reactor Engineer, Materials 9-/5- f C date-and P/ocesses Section, EB, DRS j

Approved By: ( T~7 0

-

David Terao, Acting Section Chief, Materials date ;

and Processes Section, Engineering Branch, DRS- '

-Inspection Summary: Inspection Report Nos. 50-317/90-17 and 50-318/90-15 l

l Areas Inspected: The inspectors performed a special inspection to review =several '

.self-identified welding deficiencies involving the Unit 2 pressurizer sleeve ;

replacement program, the clogging of the salt water cooling system, the damaged i LPSI pipe supports, and the licensee's evaluation of fatigue usage factors regarding the remaining life of critical component .

i 9009100149 900827 PDR ADOCK 05000317(

3 '

O PNU ;;

• <

. , ,

.

..

.

.. .

,

&

Results: The licensee's initial actions to investigate and correct the reporte(. welding deficiencies were considered satisfactory. Their initial actions to correct the_ clogging of the salt water system should improve the problem, but further testing and observation will be required to assure the system's reliability. The licensee's assessment of'the remaining life o critical components was considered technically sound. None of the critical ,

components were reported to have exceeded their usefo'. fatigue life The r- licensee's short term actions to correct the problen, of the failed../SI- '

suppt + "..hich was attributed ta check valve quick closures included analysis, operat1,; (flow) changes, and replacement of newly designed supports. Future plans include installation of new check valves which are designed to mitigate the quick closures proble :

l -

_

'

. . .

5 / 's .

. '

. ,

t i d

i DETAILS ,

- 1.0 Persons Contacted Baltimore Gas & Electric Company-

• K. R.'Boone, Project Manager

• D. P. Butler, Supervisor, Quality Control  ;

• T. J. Camilleri, Maintenance Superintendent-C.' H. Cruse, Manager, Nuclear Engineering Service' Division

!

• R. L. Demers, 0.U. Modifications Supervisor

• G. L. Detter, Director, Nuclear Regulatory Matters .

B. W. Doroshuk, Performance Engineer,. Life Cycle Management B. Dyer, Design Engineer J-

• R. A. Gambrill, Project Manager *

D. V. Graf.-P8 Project'l  ;

• K.'M. Hofiman, Supervisor, Nuclear Materials Engineerin '

• ~D. Kennedy, Mechanical Modifications-

• L. S. Larragoite, Compliance Engineer . ~

• R. M. Lloyd, Contract Administrator  !-

• M.'G. Polak, Principal Engineer, PSLO

-

• T. N. Pritchett, GS - QA A. Reed, Materials Engineering ,

• B. C. Rudell, Pressurizer Project Engineer i

• C. D. Sly, Compliance Engineer-

• J. R Sponsel, Mechanical Engineering Unit-

• E. R. Zumwalt, PE - PMU United States Nuclear Regulatory Commission (USNRC)

>

• T. J. Kim, NRC Region I, Resident Inspector.

• Denotes those attending the exit meeting on June 22, 1990. ,

l The inspectors also contacted other administrative and technical personnel during the inspectio .0 Scope of Inspection 3.0 Welding i

The inspector reviewed three we? dirty deficiencies involving the installation- Tir of' instrumentation lines as par * of the current Unit 2 heater sleev ;

replacement program. The (Miciencies were found by the licensee's project-manager and reported w the NRC site resident on: June -29,1990. During

_

the period from June 11 through June 13, 1990, the inspector had previously reviewed the major portion of the heater replacement program (sleeve / heater /

vessel connections) at performed by Batyto::k & Wilcox (B&W) including -quality control. functior, and found on deficiencies (see Inspection Report 90-13).

. _

- -

V- .

.

..

.

.:.-. ....

'

1 The socket welding in the' instrumentation lines was performed.by United Engineers & Constructors (UE&C) Catalytic, and the _ liquid penetrant and dimensional inspection activities were performed by the licensee's quality control (QC) grou The three welding deficiencies which involved the manui tungsten-inert gas (TIG) process, were identified as follows: (a) welder' performance - j qualifications; (b) limited accessibility welder qualifications; and (c) '

undersized fillet welds. With regard to welder qualification, no-deficiencies actually occurred. In this instance, the problem stemmed-from the fact that the essential information had not been entered into the computer system promptly to provide,up-to-date data-in accordance with Article QW-322'of the ASME Boiler & Pressure Vessel Code (ASME Code)Section I +

With regard to welder qual.ification for limited accessibility, the -licensee '

+

followed the guidelines provided in Regulatory-Guide 1,71 which specify a special- qualification when welder's access to a production weld is less than 12 to 14 inches in any direction from the joint. In those cases where'

accessibility for welding the instrumentation lines was judged to be marginal i by the licensee from the standpoint of the 12-inch criterion, five welders' l ,

who had made these welds were-required to requalify under restricted

. conditions. The licensee reported that the welders in question successfully welded Section IX test assemblies and provided the inspector wi h the appropriate test report .

!

With regard to the undersized welds,'the licensee reported that ofl26 welds representing fillet sizes ( 125", .195", .224"'and .312"), 16 were foun _

to be rejectable because they did not meet the minimum-leg requirement. . '

The licensee reported that additional weld' metal (portions of one layer)'.

was deposited to correct the undersized welds followed by liquid penetrant inspection Only 2 of 12 QC inspectors (weld examiners) were responsible ,

for overlooking the undersized welds. In both cases, the QC. examiners

_

l .

'

chose to use a straight edge ruler to measure the leg length even though

'

they had been trained to use fillet weld gages--the more accurate and preferred metho The NRC inspector interviewed several of the parties involved;in the weld issues including the two aforementioned licensee's weld examiners, VE&C/ Catalytic welding suptrvisor, and the licensee's~QC supervisor. From these-interviews, the irspector concluded that the undersized weld issue was an isolated incident with only two of 12 weld examiners having used-a ruler instead of gages. In one case, the examiner admitted to. complacency; in the other case, the examiner was confident of.his ability to use a rule The licensee reported that reinspection of previous welds measured by. the aforementioned examiners revealed no deficiencies. In addition, the licensee reported that a sampling of the heater sleeve / heater fillet welds, as welded and inspected by B&W (using gages), was performed with no deficiencies; foun L e

c

.

. ..-. a

. ,

... '

,

t

a

.

The problem with the undersized' welds was not picked up by the licepsee's surveillance activities even though their checklist provided an optional check point for measurement. A review of the Visual Inspection Procedure NDE 5.700 did not include instructions regarding the use of. gages for ,

measuring fillet welds:even though their training program utilized EPRI's instructions which called for gages when measuring fillet welds. The licensee reported that the problem of the undersized welds had been thoroughly reviewed with both examiners. In discussing the problem with <

.the UE&C/ Catalytic supervision, it was determined that the welders were not contractually required to dimensionally check their weld It is-an inherent responsibility of the welder, as it is, to inspect his weld for-defect Final dimensional- inspection, however, was still the- ulti!nate -

responsibility of the-Quality Control Sectio '

'

.

The inspector reinspected two of the corrected undersize welds in the, presence of one of the weld examiners who had used a' ruler instead of a gag The inspector' verified that the examiner was' proficient-in.using the-gag '

.The issues with welder qualifications, welder limited accessibility, and j undersized welds were in the process of being reviewed by the license Proposed actions under consideration included increasingtpre welding reviews with appropriate checklists,. changing; visual inspection procedure requi.re-ments, establishing mandatory s'urveillance check pr 'acs which would require welders to check completed welds with gages, and training weld examiners.' '

It should also be'noted that the intent of Regul. ,ry Guide 1.71 is for the 12-inch criterion to be used as a guideline, and that historically '

special qualification was left to the discretion of the welding supervisors or welder .0 Emergency Core Cooling System (ECCS) Salt Water Coolant System .

l -The inspector reviewed the progress of modifications to,the salt' water ,

L inlet piping for the ECCS System (ECR 85-85) discussed in Inspection Report- '

50-317/90-12 and 50-318/90-11. In this' report, it was indicated that the i modifications would be made by the licensee to the salt water piping-system branch to the ECCS pump rcom air coolers. These modifications are required

'

j l

"

because the inlet' water strainers located upstream from the air coolers tend to become clogged with crustations from the-salt water. This could; prevent the air coolers from functioning during a' loss-of-coolant-accident !

(LOCA), thereby allowing the pump room temperature to reach a'leve1~which would prevent the proper functioning of the safety injection system pumps during LOC The salt water system includes a traveling screen at the bay water intake l which is intended tc screen out the larger debris and'crustations before ;

they can enter the piping system. There is also a large concrete sump at the inlet before the talt water system pumps which contain the debris and-crustation buildup. In the sumps there is also a continuing buildup of

,

. - . . .

.

a -.

..: ..

r

.

E sea crustations on the wall that must be cleaned out on a regular. basi The' inspectors were shown a section of-pipe which had been removed from

.

the se'rvice' system because of the buildup. The buildup was readily apparent and appeared'to be dense and extremely' adherent to the pip ,

The licensee indicated that'thel smaller shells are able to pass through-the -sump and traveling screen and gain er, trance to the 8-inch header pip The take-off for the pipe 1 feeding- the ECCS system air coolers ,was located - * '

'at the bottom of the header pipe.and, as a consequence, the debris would-readily flow into the feeder line and buildup r. the strainers before:the ,

air cooler tube .

The strainers are regularly cleaned, but the possibility exists'that>

incidents of sudden influx of-debris to the feeder;lir.e could clog the screens in a short time. During an accident. requiring actuation of the pumps, the radiation level could be high, preventing any manual cleaning of the strainers forir long as seven hours. Therefore, the pump room temperature could rise to a level-preventing satisfactory operation of the ,

safety-related equipment in that room during a period of time that' the .i safety. equipment is _ needed, 4 It is essential that the strainers remain in a condition allowing flow af the cooling water during the seven-hour period that=the strainers are-inaccessible because of the radiation-levels. -The licensee intends to measure the pressure drop across the _ trainer. The pressure drop will provide an indication-of the extent of the buildup in the strainer and establish when the strainer must be cleaned out in order to survive the s seven-hour period of inaccessibility. However, the licensee'also stated that a . sudden buildup clogging ~ the screen -could _ happen sporadically and ,

concurrently with a LOC In view of the foregoing, the prevention of a sudden buildup of debris in l- the strainer is essential. The licensee's1 strategy toward that _end appears to be to prevent the debris from entering the feeder line to the strainer by relocating the 8-inch feeder line at a 90' degree angle from 'the bottom of two 30-inch' header lines. The inspector reviewed the documentation covering the relocation of the 8-inch feeder line in 122 header. The ,

records showed that the 8-inch carbon steel connection was welded.with an ASME Code Section IX qualified procedure-(PI-C/LH-AW) utilizing m double .

>

welded full. penetration joint and a fillet weld for the adjoining reinforcing plate. The records also showed that the specified lots of electrodes (E6010-and E7018) were used as well as visual' inspection reports of the completed welds. Although this modification will reduce.the possibility: of the debris '

entering the feeder line, there is still a possibility that a build up of ;

debris could occur at'the take off with a subsequent' hydraulic disturbance sweeping-a mass of debris into the feeder lin ,

i i

r i & =

.

.

. . .

,. ..-

.

,

The licensee.will also improve the salt water inlet traveling screep system ,

with a new screen system which has been found to' be more effective (two of -

-the new~ designs have already been installed). Also, the concrete inle sump w C1:be better' maintained through regular _and more. frequent _ cleanin Additionally, the chlorination system will be improved to satisfactoril deter live sea growth without causing excessive chlorine contamination o the bay.,

There are presently long . range testing and _ observation programs under way to monitor the adequacy of the planned steps to solve t'ne strainer clogging proble In view of the uncertainties expressed by the inspectors, the-licensee has committed to provide the NRC with. a discussion of its strategy in preventing the= clogging ~of the~ salt water system _ in' the near, intermediate, and long-range term. This ite.9 is unresolved pending completion of the near term plans for. resolution of the biofouling problem and review by the a NRC (317/90-17-01, 318/90-15-01).

5.0 Low Pressure Safety Injection (LPSI)-and Component Cooling Water (CCW)

'

i System Piping j ,

In Mar' '89, a walkdown of the Unit 1 LPSI' system piping by'the

-

.  !

licensee ealed a damaged piping restraint (R-2). The pipe restraint was locates in the Emergency Core Cooling System-(ECCS) pump room No. 11 and was intended to limit LPSI suction piping axial motio The LPSI system includes two pumps in parallel, with a common suction and discharge header. Each pump has a swing check valve in its vertical ,

!

oischarge line to the common discharge header. The licensee determined-that_when one pump is turned off during h.igh flow operation, a pressure transient resulting from check valve closure, (preventing flow'from the

.

operating pump from entering the discharge of the shut-down pump) causes high axial, forces on the pipe supports in the shut-down' pipin The licensee retained an engineering consulting firm to evaluate the root cause of-the-damaged piping restraint (R-2). The consultant also analyzed j- the forces throughout both the LPSI and CCW systems-(the CCW system has

'

three pumps and three check valves in parallel). Check valve quick closure (slam) tests were also conducted by the licensee to determine the loadings

,

on the system resulting from check valve slam due to the pump shifting.

!

l- The licensee concluded that check valve slam transients due to pump shifting-

' in the LPSI and CCW systems result in large . transient pressure loadings on several pipe supports. Four pipe supports in the LPSI system were damaged:

R-2 in Unit 1 suction line, R-16 and H-18/R/15 in the Unit 1 discharge line, and R-16 in the Unit 2 discharge line.

,

I.

'

r

~

.- .

-

. .. ,

,

.

.. .

,

Testing and analyses have been used by the licensee and their consultan to determine the loads on the supports of the LPSI and CCW piping systems-due to check valve slam. An evaluation of the. pipe support' compared to:

the calcula+ed loads-produced results that are consistent with the. damage-reported for R-16, the damage to the original'. design of Unit 1 R-2, and are consistent with the ad tquate performance of the. currentidesign o Unit 1 R-2. The anaiiras also. indicate that the check valve sla- ' ads a many.other su'pports-exceed their' design capacity although~they have not, '

as yet, shown signs of darrag The licensee determined, as a result of testing, that lesser check valve a slam forces occur dering pump shift if the flows are reduced to 900 GPM in the LPSI system (below the level of slams observed.for pump shift at:1500 to-3000.GPM). Hence, the licensee believes mitigation of check valve slam could be achieved through flow reduction to .800 GPM during pump shift.- .l The licensee conta'cted ti check valve manufacturer (Mannesmann Demag) who could provide'a check valve design utilized in European.. power plants which has been shown to elir.inate check valve slam during normal operating flow i pump shifts. The rapid, shock-free closing' operation'of this valve is .f' 1 manifested through the small mass of its moving parts, adjustability of its springs, and a short operating stroke (in contrast to the higher mass and long stroke of the conventional swing check valve). The licensee has committed to changing the check valves to the Mannesmann design'as a solution to the pump shift check valve slam problem. Until the replacement is complete, however, the licensee will change its operating procedure during the pump shift operation to reduce flow. Furthermore,.the pipe support system will be repaired and, where necessary, strengthened'through. redesig BG&E has provided for an engineered solution to the pump shift check'

?

valve slam problem through a program ~of analysis, testing, replacement of a newly-designed supports, and review of optional check valve designs. The problem solution provides for an interim mitigation of the check valv slam forces and final elimination of these forces through replacement of the existing check valves with those of an improved desig .0 Critical Component Fatigue Life Usage Assessment The design of components for the nuclear steam supply systems at Calvert Cliffs Units 1&2 was performed'in accordance.with the rules of.Section III of the ASME Boiler and Pressure Vessel Code - Nuclear Components. In the design rules of this Code,-it is necessary for the designer of the components to provide for analyses of the components subjected to the cyclic thermal and mechanical loadings over a predicted life span expressed in the equipment i specifications for the plant components. The analyses, therefore, were-mostly directed toward.the stress / strain cycling resulting from the operating cycles of the plant, such as startup, shutdown, load changes, upset and s emergency conditions. Resulting from the analyses are estimates of the i

!

a

'

- .,

-

. >

.e .

cumulative fatigue life usage factors for criticali sections of the' component l For an individual loading, the usage is the actual-predicted number o cycles for a stress / strain range for.the loading divided by:the maximum ,

'

number of cycles that could be achieved without fatigue failure for1that'

stress / strain rang t

~

In order for the -licensee to ascertain whether it has exceeded.the useful '

life of-the component understhe actual loading conditions on the component '

throughout its life to date, a record-of the actual numbers of cycles ,

achieved-to date since the start of; plant operation is required. Together-with:the stress analyses of the components at its critical' areas, the licensee must ascertain that the cumulative usage-factor:at.each critical area does not exceed the, allowable'value of The licensee discussed with the inspector.the system utilized by BG&E at 1 Calvert Cliffs 1&2 to monitor the fatigue usage of the plant equipmen 'The inspector observed a written log.taken of plant operational cycles-kept in the control, room as part of the. normal plant monitoring procedures. At . regular intervals, this log _is reviewed by engineering ;

personnel and compared with the ' predicted numbers of cycles for the - J component that formed the basis for its design analysis. This is a quality control procedure to provide for cbservation of the fatigue usage of each critical componen The engineering overview of critical plant component fatigue; usage is the responsibility of.a performance engineer in life-cycle management. Based on a design basis' defined in equipment specifications ~,JASME Section III analyses and design reports, and B31.1 design: reports the design envelope-

,

is reflected in the plant technical specification safety-analyses report, and licensing commitment TheLfatigue. status of the Calvert Cliffs 1&2 Station-is determi%d through identification of safety related systems, key components 'in each system,. definition 'of significant loads, stress cycles for each load type and usage. factor calculations.

L As a consequence of the review of life cycle management procedures at Calvert Cliffs 1&2 as explained by the licensee performance engineer, there is evidence of a strong activity in the evaluation of. the remaining life of critical components for fatigue usage. The approach is' consistent with the development of remaining life assessment technology. The licensee reported that, thus far, none of the' critical components have exceeded their useful f atigue life'l The inspector found.no deviations'or violation in this area and believes this activity at BG&E Calvert Clif fs 1&2 to be satisfactory.

L l

l t

I

.

'-

.'-,

.

,

.. .

-10

.

7.0 Conclution ,

The licensee's initial actions to investigate.and correct the reported welding deficiencies is considered satisfactory. The. licensee has been <

requested to submit to the NRC their short-term corrective actions to .l prevent rocurrenc The licensee's initial actions should improve the'. l problem of tha clogging of the salt water system, but further testing and l observation will be required to assure the system's reliability under' ;

extreme conditions. The-licensee has been requested to. submit its strategy to prevent clogging of-the salt water system to-the NRC.- The. licensee's approach to deterr.ine the remaining life of the critical components is ..!

considered technically. sound. None of the components were reported.to, I have . exceeded their useful fatigue life. The licensee's short-term actions to correct the cause of the-damaged LPSI pipe supports which was attributed to check valve quick closures included analysis, operation (flow)' changes,z and replacement of newly designed supports. Future plans include installa-tion of.new check valves designed .to mitigate valve quick closures'.

An exit meeting was conducted on July 13,1990 (see paragraph 1.0 for, ,

attendees) at which time the findings of the inspection were presente '

No written material was provided to the licensee by the inspector, i

i

.d

!

$

i

,. l i

a 1