ML20197C454

Safety Evaluation Supporting Resolution of Nine Mile Point Reactor/Turbine Building Pressure Relief Panel Outside Design Basis Issue

ML20197C454
Person / Time
Site:	Nine Mile Point
Issue date:	01/02/1997
From:	NRC (Affiliation Not Assigned)
To:
Shared Package
ML20197C458	List: IA-97-375, Forwards SE Re Reactor & Turbine Building blow-out Panels ML20197C454 ML20197C477 ML20203J447 ML20203J461 ML20203J482
References
FOIA-97-375 NUDOCS 9712240191
Download: ML20197C454 (3)

Text

-_

e e Mog

' y *- 4 UNITEG) STATES j j

} NUCLEAR REGULATORY COMMISSION i

• & WA&HINGTON, D.C. enm mg l SAFETY EVALUATION SY THE OFFICE OF NU LEAR REACTOR REGULATION

'*U*****.! RESOLUTION OF NINE MILE POINT 1 kEACTOR/ TURBINE BUILDING PRESSURE REllEF PANEL DUTSIDE THE DESIGN BASIS ISSUE NIAGARA M0 HAWK POWER CORPORATION NINE MILE Pn!NT UNIT 1 DOCKET NUMBER 50-220 1.0

• 1HI@pVCTION On March 13, 1996, Region I (RI) requested assistance from the Office of Nuclear Reactor Regulation to determine the technical adequacy of Niagara Mohawk Power Cnrporation's (NMPC) revised design calculations for the reat. tor (TB) blowout panels. These calculations building are related(RB) and to an turbine event (LERbuilding 50-220 /95-05) that identified the existence of an initial construction deficiency and a subsequent 1993 error in design assumptions for the blowout panels. The staff reviewed NMPC's revised design calculations and attended an NRC enforcement conference between RI staff and NMPC held on April 12, 1996. A set of technical questions generated in the staff reviev. was forwarded to NMPC as an attachment to NRC Special Inspection Report 50-220/96-05; 50-410/96-05, dated March 29, 1996. Two tele) hone conversations between the NRC staff and NMPC representatives were leid on May 2 and May 22, 1996, to discuss engineering analyses being implemented by the licensee for the resolution of the blowout panel issue. NMPC responded to the staff technical questions on July 3, 1996. The staff conducted an audit of the licensee's reactor end turbine building blowout panel calculations at the plant site on August 20, 1996. This led to identification of additional items requiring licensee response and resolution. The staff reviewd the licensee's response and disposition of the items submitted on November 15, 1996. The following evaluation provides the basis for resolution of Nine Mile Point 1 RB/TB pressure relief panel design basis issue.

2.0 EVALUATION Based on the NRC staff's technical questions and discussions held with the staff, NMPC decided to re-analyze the upper-bound pressure capacities of the reactor and turbine building pressure relief panels and the lower-bound structural failure capacities of their respective buildings to demonstrate that the panels will fail due to internal pressurization prior to failure of the buildings, thus, achieving the intended pressure relief function stipulated in the Nine Mile Point Unit 1 Final Safety Analysis Report (FSAR).

The RB blowout panel bays measure 19 feet in width and 44 feet in height.

Each bay consists of 24-inch wide horizontal panels of metal decking, bolted to vertical building columns and interlocked vertically by crimped side joints. The bays in the RB are also bolted at the top and bottom. The bays of the TB measure 20 feet by 20 feet, and consist of panels that are similar to those of the RB. However, the bays in the TB are not bolted at the top.

9712240191 971212 ELIAh-375 PDR ATTACHMENT ,

,p p M O NI

_ _ - _ _ _ - - _ . - . . .- _ - _ - - = - _ _ _ . .. -

i in the analyses of the upper-bound panel blowout capacities for the reactor building and turbine building, NMPC adopted assumptions including: the shear bolt failure determines panel blow-out capacity; the 24" wide panels span horizontally with one way action; the tens ~ile capacity of bolts is equal to the highest value obtained from the RB and TB bolt sample tests; the shear capacity of a bolt is approximately 0.62 of the tensile capacity; the elliptic shear tension interaction equation adequately defines the bolt failure; the .

use of the actually installed bolt size and spacing information established '

through plant walkdown; the consideration of dead weight effect; and the incorporation of the minor effect of large-displacements and boundary angle deformation on the bolt tension. The panel upper-bound blow-out capacities determined from the analyses based on the above assumptions are 65 psf and 62 psf for the RB and TB, respectively.

The staff reviewed the calculation of the largest blowout pressure of the panels fer both buildings. Based on the plant walkdown and assessment of panol configurations, we agree with the licensee's assumption that shear bolt failure determines panel blowout capacity. The licensee stated that the appropriateness cf its assumption of one-way span action for the panels was confirmed by a specific evaluation of the effect of membrane action via a finite element analysis using the ADINA computer program. We have reviewed the specific evaluation and concluded that the one-way assumption is reasonable and acceptable. The licensee's use of hignest tested bolt tensile capacity and 0.62 of the highest tensile capacity as bolt sheir capacity as well as its use of an elliptic shear-tension interaction equation in analyzing the panel blowout pressure capacity were also reviewed for its adequacy, and were found acceptable. The method used in accounting for the dead load, large-displacement and boundary angle deformation effects was also assessed and found acceptable. In st eary, the staff agrees with the method used by the licensee in determining ' W blowout panel capacity and considers the 65 psf and 62 psf upper-bound blowout panel capacities established for the RB and TB, respectively, acceptable.

To analyze the minimum failure capacities of the RB and TB building super-structures, the licensee used detailed COSMOS models to represent the structures and applied dead load and a reference internal pressure load of 93 psf. In the COSMOS model, the 24" wide FKX steel siding is considered to provide continuous lateral-torsional support to column flanges of the structures. The licensee used yielding strength based upon minimum values from available alli certs in evaluating the lower-bound structural capacities.

The analysis indicated that the pressure capacities for the structures were controlled by a roof bracing and a critical vertical column. The AISC Steel Construction Manual was used to determine the lower bound capacities of 143 psf and 135 psf for the RB and TB buildings, respectively. In evaluating these analysis results, the NRC staff found that a linear scaling method which /

the licensee used to obtain the pressure capacities from the COSMOS computer analysis results needed additional adjustments in order to provide correrg -)(

results. The licensee implemented the needed adjustments including performance of an independent dead load case analysis for the RB and TB using the COSMOS model and application of a 0.9 capacity reduction factor to the COSMOS analysis results. The licensee also agreed to eliminate the 2

I consideration of strain rate effect due to rapid load appilcation in /

establishing the Fy value used in the structural capacity evaluation. The adjustments resulted in revised lower-bound failure capacities of 117 psf and ',

135 psf for the RB and TB, respectively. These capacities are shown to be much grer.ter than the maximum computed blowout panel capacities of 65 )sf (RB) and 62 psf (TB), and thus, ensure timely pressure relief function of tie  ;

panels prior to RB or TB structural failure. Therefore, these lower-bound  ;

structural failure capacities are acceptable to the NRC staff.

On August 20. 1996, as part of a plant design audit conducted at the site, the staff performed a walkdown of the RB and TB and examined the modified .

pressure relief panels including the increased shear bolt spacings and the installed conditions of the 24-inch FKX panels. The staff found that the blowout panels for the two buildings were in excellent condition and concluded that there is a reasonable assurance that the panels will perform their intended pressure relief t 4ety function during a design basis accident as stipulated in the FSAR.

3.0 CONCLUSION

Based on our evaluation of licensee submittals, discussions held with NMPC representatives, audit of engineering calculations and a plant walk down of the pressure relief blowout panels performed at the plant, we concluded that the panels will perform their intended pressure relief safety function in a

".nner consistent with the applicable licensing bases sti)ulated in the plant

.5AR, thus, the panels as modified and supported by the a)ove discussed engineering analysis results are acceptable. This conclusion should suffice as basis for closure of the Nine Mile Point I reactor and turbine building blowout panel of design basis issue.

3

,