IR 05000298/1986030

Insp Rept 50-298/86-30 on 861103-07.Violation Noted:Failure to Measure Individual Leak Rates Prior to Making Repairs

ML20215D502
Person / Time
Site:	Cooper
Issue date:	12/01/1986
From:	Hunnicutt D, Jaudon J, Mcneill W, Tapia J NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV)
To:
Shared Package
ML20215D467	List: IR 05000298/1986030 ML20215D464 ML20215D484
References
50-298-86-30, NUDOCS 8612160383
Download: ML20215D502 (10)

Text

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

!

.. ,

APPENDIX B U.S. NUCLEAR REGULATORY COMISSION

REGION IV

NRC Inspection Report: 50-298/86-30 License: DPR-46 Docket: 50-298 Licensee: Nebraska Public Power District P. O. Box 499 Columbus, Nebraska 68601 Facility Name: Cooper Nuclear Station (CNS)

Inspection At: CNS Site, Brownville, Nebraska and General Offices (GO),

Columbus, Nebraska Inspection Conducted: November 3-7, 1986 Inspector: *

W. M. McNeill, Project Engineer, Project M[ /1/'/df-Date Section A, Reactor Projects Branch (pars. 1, 2, 3, and 6)

bhl) OnN J. I. Tapia, Reactor Inspector, Operations

/d///84 Dat'e /

I Section, Reactor Safety Branch (pars. 1, , 4, 5 and 6)

Approved: th/ _M /A J. Pf. Jauf n, Chief, Project Section A Date JeadtorPtrojectsBranch h)

D. M. Hunnicutt, Chief, Operations Section

>> W /A///W6 Dat'e '

Reactor Safety Branch Inspection Summary Inspection Conducted November 3-7, 1986, (Report 50-298/86-30)

Areas Inspected: Routine, unannounced inspection of onsite followup of written reports of two nonroutine events and actions on previously identified inspection finding Results: Within the areas inspected, one violation was identified (failure to measure individual leak rates prior to making repairs).

h G

bo J

[

. - - _ - - - - _ _ _ _ _ _ _ _ _ _ _ _ _ - _ - _ _ _ _ _ _ _ _ _ - - _ _ _ - _ - _ _ _ - _ - - - _ _ - _ - _ _ _ _ - _ - _ _ _ - _ - _ _ _ _ _ _ -

s .

, .

. ..

-

-2 -

.

-

DETAILS '

. Persons Contacted NPPD

, *R. Brungardt, CNS Operations Manager

,

P. Chuansnit, Cooper Industries, Field Service Representative W. E. Crawford, CNS, Maintenance Supervisor

• K. J. Done, G0, Engineering Supervisor R. Gibson, CNS, QA Specialist

• A. P. Heymer, G0, Engineer J. D. Hilgenkamp, CNS, Mechanical Engineer D. A. Hopper, CNS, Mechanical Foreman

'G. R. Horn, CNS, Division Manager Nuclear Operations W. Horstman, Cygna, Engineer

• E. M. Mace, CNS, Engineering Manager

, *G. S. McClure, GO, Nuclear Engineering Department Manager

'J. M. Meacham, CNS, Technical Staff Manager

'C.' R. Moeller, CNS, Technical ~ Staff Supervisor

'D. M. Nowell, CNS, Maintenance Manager

• L. Shipley, Cygna, Regional Office Manager W. J. Siske, CNS, Mechanic

• G. E. Smith, CNS, Acting QA Manager M. E. Unruh, CNS, Maintenance Planner / Scheduler

• N. Williams, Cygna, Manager Engineering Mechanics Division R. L. Yuntz, G0, Civil / Structural Enginee * Denotes personnel attending exit meeting at General Offices, Columbu * Denotes personnel attending exit meeting at CNS sit The NRC inspectors also contacted other personnel including administrative and clerical personne . Licensee Action on Previously Identified Inspection Findings t

(Closed) Unresolved Item (298/8530-02): QA qualification records were not available to demonstrate what background data was used to meet the equivalent requirements in lieu of degree The licensee has issued a memo to the file for all QA specialists and engineers detailing the background data used to meet the equivalent requirements in lieu of degrees. The records of five personnel were reviewed and found to be satisfactor This item is close o .  ;

,

.- .

A-3-(Closed) Open Item (298/8008-01) Review of Cooper Nuclear Station Summary Technical Reports on Primary Containment Integrated Leak Rate Test (ILRT).

During the onsite inspection of the second periodic Containment Integrated Leakage Rate Test (ILRT), a valve was found to be leaking in excess of the previously determined type C test results. The licensee repaired the valve stem packing and committed to provide information on the corrective actions taken in accordance with Appendix J,Section III.A.1. Since the NRC inspector did not address this matter during the conduct of the type A test, an open item was documented in NRC Inspection Report No. 50-298/80-0 The NRC inspector reviewed the final reports for the second and third periodic ILRTs submitted pursuant to the requirements of 10 CFR Part 50, Appendix J. The report for the second periodic ILRT was found to violate the leakage testing requirements of Appendix J,Section III.A.1(a).

This section requires that, "If during'a Type A test, including the supplemental test specified in III.A.3(b), potentially excessive leakage paths are identified which will interfere with satisfactory completion of the test, or which result in the Type A test not meeting the acceptance criteria III.A.4(b) or III.A.5(b), the Type A test shall be terminated and the leakage through such paths shall be measured using local leakage testing methods. Repairr, and/or adjustments to equipment shall be made and a Type A test performed. The corrective action taken and the change in leakage rate determined from the tests and overall integrated leakage determined from the local leak and Type A tests shall be included in the report submitted to the Commission as specified in V. B."

Contrary to these requirements, it was determined from a review of the submitted report that valves HPCI-M0-19 and RCIC-37 had been repaired after excessive leakage was identified and that the leakage rate prior to repairing the valves was not measured or documente The "as found" condition of the containment was, therefore, not able to be define This is an apparent violation (298/8630-01).

This open item is closed and further action will be tracked as a violatio . Standby Diesel Generator During the 1986 annual inspection of diesel gererator No. 2, Surveillance Procedure 6.3.12.6, a borescope examination was performed on all 16 cylinders. Three cylinders had their heads removed because the cylinder liners were suspected of being scored. The visual examination that followed established that the cylinder liners were acceptable; however, the cylinder heads appeared to be cracked in the fire dome are Additional heads were removed and found to be in a similar condition. All 16 cylinder heads were removed and shipped to Tulsa, Oklahoma, to the Cooper Industries Service Center. These were inturn shipped to a subvendor, Reynold & French Company, also of Tuls At Reynold & French

!

I

e _

.

.

-4-Company the heads'were subjected to magnetic particle inspection and

. hydrostatic testing. It was found that, of the 16 cylinder heads, four had no evidence of cracking, eight had cracking, but were considered e useable, and four were rejected. Of the four that were rejected by the Cooper Industries personnel, two were rejected because they had failed the hydrostatic test and two were rejected because of cracking and excessive pittin Licensee Nonconformance Report No. 005189 was written to document the problem, and a licensee event report (No. 26) was issued by CN Background The standby A. C. power system at CNS consists of two independent diesel generators which' automatically start and supply power in the event of loss of offsite power to satisfy emergency loads. Two standard, commercial KSV-16-T model diesel generators were built in 1970 by Cooper-Bessemer Reciprocating of Cooper Industries. The diesels are four-cycle V-16 engines which rated for 5560 horse power at 600 rpm. There are two cylinder blocks, one for each eight cylinders, which are mated to individual cylinder heads. The cylinder heads contains two intake and two exhaust valves, one fuel injection nozzle and a starting air valve. Each head is cooled with jacket water from the cylinder block. The cylinder heads are made of high strength gray iron (Mechanite) castings which are stress relieved. The number two diesel was put into service in 1973 and has had about 760 starts and 1200 running hcurs. These diesel heads at CNS have a gas / diesel configuration. This means there are additional passages in the head for gas operation in lieu of diesel fuel. It was noted that these additional passages sacrifice some cooling volume. The diesels are subject to two types of survie11ance tests as required by Technical Specifications (paragraph 4.9.A.2). At least monthly each diesel is subjected to a four-hour run at 80 percent load per Surviellance Procedure No. 6.3.12.1 Also there is an annual inspection per Surveillance Procedure No. 6.3.12.6, which incorporates the vendors recommendations as found in the vendor technical manual and service news item Corrective Actions CNS response to the cylinder head cracking was to ship the heads to Tulsa as noted earlier. It was noted by the NRC inspector that the magnetic particle inspection and hydrostatic testing performed in Tulsa by Reynold

& French Company was incompletely documented. Comparison of testing done in Tulsa to that performed by Cooper Industries during manufacturing of nuclear grade heads indicated significant differences. The magnetic particle testing done in Tulsa was by the yoke technique, whereas in manufacturing the coil technique was used. Because two different techniques were used, it was difficult to establish the magnetizing force adequacy in terms of direction and strength. The hydrostatic tests in Tulsa was performed for one hour at greater than 90 psig and 170 F. In manufacturing hydrostatic testing is performed for ten minutes at greater than 100 psig at room temperature. More importantly, testing in Tulsa did

._ ,- - --

.

.

-5- .

not isolate the pressure source nor vent all air from the system teste In manufacturing such is done. In summary, the sensitivity of the testing !

, performed in Tulsa is unknown. Photographs were taken of the crack indications at Tulsa. The four heads that were not cracked were IL, SL,

'

-

8L and 1R. The eight heads that were cracked.but considered acceptable were 3L, 6L, 7L, 2R, 3R, SR, 7R, and 8R. The~ heads that were leakers were 4L and 4R. The heads that were cracked and pitted and also rejected were 2L and 6R. The pattern of cracking was invariably in the web area between either exhaust and intake valves or between exhaust valves and the air start valve. CNS QA personnel performed surviellance of Reynold & French Company activities and of the relapping and blueing performed at Cooper Industries, Tuls It was noted the exhaust and intake valves were relapped and blued at the same time as the cracks were studied. CNS issued a purchase order (P. O. 261602) to Cooper Industries for replacement heads. Cooper Industries, Grove City, Pennsylvania, was audited by G0 QA personnel. The receipt inspection records were reviewed by the NRC inspector. It was noted that the surveillance report of Grove City concluded that only difference between nuclear grade heads and commerical heads was the paper work (i.e., traceablity, certifications, etc). However, in review of the certifications for the new heads and the QC plan for nuclear grade heads it was noted by the NRC inspector that the heads were not magnetic particle tested. The new heads were commerical heads being upgraded by CNS. Only about five heads were available for use by CNS in the Grove City manufacturing process. These new heads were a diesel only configuration. As new diesel configuration heads are available from Grove City, they will be used to replace the diesel / gas configuration heads on number two diesel. This is documented in Engineering Specification Change No. 86-87. The time table has not been established for the replacement of the remaining twelve heads. As soon as diesel number two is operable with the four replacement heads, then diesel number one will be inspected for cracking of the cylinder heads. A Special Procedure No. 86-16 has been draf ted to attempt to verify water intrusion into the cylinders during the operation of number two diesel with the eight heads known to be cracke Conclusions Cooper Industries has documented in a memo to CNS dated October 29, 1986, that the cylinder heads are subjected to thermal stresses. The thermal stresses are a function of high peak pressures, and temperature cooling water temperature, air manifold temperature, etc. The thermal stresses of a diesel / gas configuration are implied to be greater than that of a diesel configuration. Cracking would be detectable shortly af ter shutdown by rolling the engine with the indicator cocks open given that the cracking had propagated to the cooling water jacket. The implied conclusion of the memo is that the cracking is the result of tensile stress during cool down from abnormal thermal cycling. It should be noted that CNS does have lube oil and jacket water heaters, which continously heat the diesel when it is not in operation. A review by the NRC inspector of the monthly surviellance reports for the past two cylinders rejected, one cylinder (2L)had years indicated a very that of of distinct pattern the four

_- _ ._ _ _

- _ _ _ _ _ _ _ _ _ _ _ _ . ._ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ - _ _ _ _ _ _ _ - _ _ _

_ _

. ..

-6-

.

running hot (greater than 100*F above the temperature of the other cylinders) during the last half of 198 The NRC inspector was told that the Zion site had found cracked cylinder heads and that their diesel / gas configuration heads had been. replaced with diesel configuration head The NRC inspector expressed concern that there was no evidence to indicate the mechanism of cracking of the cylinder heads. There was also no evidence to indicate if the cracking at CNS was at all related to the cracking experience at the Zion site. Cooper -Industries indicated during a phone conversation with the licensee that they may have some evidence to demonstrate how and when this kind of cracking could occur. The NRC inspector was concerned that the adequacy of the corrective actions can only be evaluated by establishing how and when this cracking of the cylinder heads occurred. This is an open item, pending evaluation of additional data (298/8630-09). Seismic Design of Heating, Ventilating and Air Conditioning (HVAC)

Systems During the week of October 20, 1986, the licensee informed the NRC that the suction side of the Standby Gas Treatment (SBGT) system did not meet the requirements of a Seismic Class I system as delineated in the Updated Safety Analysis Report (USAR). The licensee reported that the installed condition of the SBGT system supports did not meet the original design specification requirements. This determination resulted from an independent reevaluation of the seismic design and as-built condition of the suction side of the SBGT System performed by Cygna Energy Services, In The Cygna evaluation was an extension of a review performed by Stone &

Webster Engineering Corporation (SWEC) in 1985. The SWEC review of the SBGT System was initiated to satisfy a commitment made by NPPD in their response to the April 4, 1985, Notice of Violation concerning a failure to maintain secondary containment integrity during refueling activities conducted on September 22-29, 1984. The commitment was to retain an independent architect / engineering (AE) firm to conduct an independent review of the SBGT system, intended to determine if modifications to the system were required and to determine if changes were necessary to improve testing procedures, operating procedures, or the Technical Specification The scope of the SWEC review was to ensure that the SBGT system was in compliance with the applicable requirements of Regulatory Guide 1.52,

" Design, Testing, and Maintenance Criteria for Post Accident Engineering-Safety-Feature Atmosphere Cleanup System Air Filtration and Adsorption Units of Light-Water-Cooled Nuclear Power Plants," and ANSI Standard N509, " Nuclear Power Plant Air Cleaning Units and Components STET," 198 The SWEC Final Report, entitled " Standby Gas Treatment System Review," and issued August 6,1985, found that a crossover duct between the two SBGT system trains was unrestrained in one direction and that documentation was not available at the site to confirm the seismic adequacy of the remaining

- - _ _ - -- _ _ - - - - - _ _ _ _ -

_ _ _ _ _ _ _ _ _ - _ _ - _ - __ _ _ _ _ _

.. . .

-7-supports in the system. SWEC recommended additional bracing on the supports for the crossover duct and verification of the seismic adequacy of all other SBGT supports by obtaining tne calculations from the original designers or by performing new calculation On July 1,1986, NPPD commenced an internal document search and requested the original AE, Burns & Roe, Inc., to also search for documents relating to the seismic adequacy of the SBGT system. Similar requests were issued to Waldinger Corporation, the HVAC installer, and to Peterson Engineering Company, the HVAC support system designe Cygna was concurrently tasked with performing an as-built walkdown of the suction side of the SBGT system. Comparison, by Cygna, of the results of their walkdown with calculations received from Peterson Engineering indicated disagreement between the design and the installed conditio Nonconformance Report NCR No. 4642 was then issued on October 21, 1986, to document the failure to meet seismic class I design criteria. The disposition of the NCR required generation of a design change, as-built drawings upon completion of the modifications and an expanded inspection to include all seismic class I HVAC installed by the some contracto CNS Station Design Change No.86-049 delinfates the modifications which were necessary to seismically qualify fif ty-eight supports on the SBGT system. The NRC inspector reviewed Cygna Document No. DC-1, Revision 1,

" Seismic Evaluation Criteria and Methodology." This document provided the basic criteria and methodology which was used for the field walkdown, the as-built seismic evaluation, and the modification of the existing syste The design analysis utilized by Cygna to qualify SBGT system supports was based on the combined deadweight and seismic loads of the SBGT System duct. The seismic evaluation was performed to ensure that the stress levels met the applicable requirements of the ASME Code,Section III, Division I, 1983 Edition. The NRC inspector determined from a review of the reevaluation criteria that the weight of the support structure itself was not considered in the analysis. The analytical basis for the omission of the support weight is considered an unresolved item (298/8630-02) which will be addressed during a subsequent inspection. The following criteria was also used by Cygna in their support evaluation:

"

Structural steel members were evaluated to the provisions of the American Institute of Stcel Construction (AISC), " Specification for the Design Fabrication, and Erection of Structural Steel for Buildings," 7th Editio Welds were checked to the provisions of AISC and the American Welding Society (AWS), " Structural Welding Code." D1.1-/ * The allowables used for the Phillips red head anchor bolts were based on self-drilling anchors installed in 5000 psi concrete. A factor of safety of 5 was applied to the vendor published ultimate tension and l

. -- - . . _ _ - - - _- -- .

p

-. .. ,

!

-8-

~

shear loads in accordance with USNRC IE Bulletin 79-02, March 8,1979,

" Pipe Support Base Plate Designs Using Concrete Expansion Anchor Bolts."

Tension and shear interaction was checked as recommended in

" State-of-the-Art Report on Steel Embedments," American Society of Civil Engineers, Nuclear Structures, and Materials Committee, June 198 The formula for checking tension and shear interaction recommended in the l last reference above is based on a plot of shear versus tension which is elliptical and not as conservative as the linear interaction which has been typically used in original designs of anchor bolts. The NRC inspector requested a copy of this reference in order to verify that the recommended formula was applicable to the Phillips Red Head anchor bolts utilized in the original construction. This infomation was not provided during this inspectionandisthereforeconsideredanunresolveditem(298/8630-03)to be addressed in a subsequent inspectio i The Cygna field walkdown provided the as-built information which was used in the seismic analysis of the suction side of the SBGT system. The as-built information generated by Cygna consists of individual duct support drawings and four duct isometric drawings (No. SBGT-S-1, Sheets 1 through4). Using the four isometric drawings, the NRC inspector performed an independent walkdown of the suction side of the SBGT system. Three dimensional discrepancies were identified. It could not be determined during this inspection whether or not the noted discrepancies had been included in the seismic analysis calculations or were draftin This matter is considered unresolved pending further review (g errors.298/8630-04).

The welds connecting the SBGT system duct to the supports are being evaluated by Cygna to the provisions of the aforementioned AISC and AWS codes. The duct is made of Type 304 stainless steel, while the supports are constructed of Type A 36 carbon steel. Cygna has assumed, "in the absence of weld documentation," that the welds are compatible with the duct materials and therefore, that, " weld material properties shall be chosen accordingly."

The original design requirements are listed in the contract between NPPD and Waldinger Corporation. Contract No. E-70-19, Section G, Part II,

" Technical Specification " requires strict conformance with the latest edition (in effect at the time of bidding, March 2, 1971) of the applicable codes, standards, and specifications of the American Society of i Mechanial Engineers (ASME), American National Standards Institute (ANSI), l AISC, and AW The code applicable to the original qualification of the support duct weld is ASME,Section IX, " Welding Qualifications," 1968 with addenda through Winter of 197 Paragraph Q-10 (a) of this standard requires each manufacturer or contractor to record in detail and qualify I the procedure specification for any welding procedure that he follows in the construction of weldments. As of this inspection, NPPD was not able

. . . .

-9-to locate the required documentation. This matter remains unresolved (298/8630-05) pending completion of the ongoing document search and receipt of additional analysis of the dissimilar metal weld which the licensee comitted to provide at the exit intervie The expanded inspection of other HVAC systems has, as of this inspection, identified the control room ventilation system as the only other HVAC system that was designed as Seismic Class I. The NRC inspector reviewed the generic support calculations documented in Peterson Engineering calculation No. CWP, July 26, 1971, " Control Room Duct Hangers Class IS."

It was determined that anchor bolt shear and tension interaction had not been checked, the weight of the support was not included, no weld calculations had been performed, and the members subject to both axial compression and bending stresses were not analyzed in accordance with Section 1.6 of the AISC, " Specification For The Design, Fabrication &

Erection of Structural Steel For Buildings (February 12,1969)." NPPD has not concluded their document search for additional calculations:

therefore, the observations made by the NRC inspector are considered an unresolveditem(298/8630-06).

No individual support detail drawings of the control room ventilation system had been located as of this inspection. The as-built condition, therefore, cannot be defined without a walkdown of the system. This effort commenced on the last day of this inspection. The results of the walkdown are considered an unresolved item (298/8630-07) which, when answered by the licensee, will define the as-built condition of the support NPPD has also commenced a document search of four other systems; the quadrant room ventilation systems, the purge / vent system, the reactor building ventilation system, and the diesel / generator ventilation syste This search has identified Peterson calculation No. CWP, August 13, 1971,

" Duct Hangers - class II S Seismic Loading -Reactor Bldg. For Table A -

Class I Galvinized Ductwork-Rectangular." These calculations were reviewed by the NRC inspector and found to exhibit the same omissions as the control room ventilation system calculation The seismic loading requirements for equipment provided under contract E-70-19 call for seismic horizontal and vertical forces of 1.5W and 0.21W for Seismic Class I, and 0.1W and 0 for Seismic Class II, where W is the weight of the item under consideration. The resultant seismic forces are utilized an static load Section XII of the facility Updated Safety Analysis Report (USAR) states that a Class II designated item shall not degrade the integrity of any item designated Class I. The NRC inspector questioned the ability of Seismic Class !! items designed to such low seismic loadings to satisfy the intent of the USAR statement. Reconciliation of this issue is considered an unresolved item (298/8630-08) which will be addressed during a future inspectio No violations or deviations were identifie . .

__ _ . _ _ _ _ _ _ . _ _ _ _ _ _ _ _ - _

_ _ _ _ _ _

• , .

-10- Unresolved Items Unresolved items are matters about which more information is required in order to ascertain whether or not the items are acceptable, violations, or deviations. The following unresolved item are discussed in this report:

Paragraph Item Subject 4 298/8630-02 Support Self Weight 4 298/8630-03 Tension / Shear Interaction 4 298/8630-04 Dimensional Errors 4 298/8630-05 Welding Procedure 4 298/8630-06 Control Room Calculations 4 298/8630-07 Control Room Walkdown 4 298/8630-08 Seismic Class II Design Exit Meetina l The NRC inspectors conducted separate exit meetings on November 7, 1986, with the licensee personnel denoted in paragraph 1. One exit was at the site with both NRC resident inspectors and project manager attending, and the other exit was at the General Office. At the meetings, the scope and-findings of the inspection were summarized.

i l

l l

l i

,