ML20078P725

Affidavit of Ew Thomas Re Revised Design Response Spectra

ML20078P725
Person / Time
Site:	Callaway
Issue date:	11/03/1983
From:	Eric Thomas BECHTEL GROUP, INC., UNION ELECTRIC CO.
To:
Shared Package
ML20078P720	List: ML20078P713 ML20078P725
References
ISSUANCES-OL, NUDOCS 8311080205
Download: ML20078P725 (8)

Text

.

. UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION BEFORE THE ATOMIC SAFETY AND LICENSING A? PEAL BOARD In the Matter of )

)

UNION ELECTRIC COMPANY ) Docket No. STN 50-483 OL

)

(Callaway Plant, Unit 1) )

AFFIDAVIT OF EUGENE W. THOMAS ,

County of Montgomery )

) ss.

State of Maryland )

EUGENE W. THOMAS, being duly sworn, deposes and says:

1. I am employed by Bechtel Power Corporation, Gaithersburg Power Division, Gaithersburg, Maryland. My present position is Civil / Structural Engineering Staff Supervisor. I have previously testified on behalf of Applicant in this proceeding on the embedded plate issue, submitting prefiled written testimony dated November 6, 1981 (Applicant Embed Testimony) and appearing for oral testimony at the hearing on this issue on November 18-23, 1981. A complete state-ment of my professional qualifications is incorporated in Applicant Embed Testimony, following Tr. 501, at 4 and Attachment 2.

2. I previously prepared an affid,avit, dated October 11, 1983, which was filed in support of (and as Exhibit D to) Applicant's Response to Joint Intervenors' Petition for Reconsideration, dated October 12, 1983. I make this Affidavit in response to the Appeal Board'.s Memorandum and Order of October 20, 1983, which seeks addi-tional information with respect to an Observation (No. 4-1) made in the Integrated Design Inspection Program (IDIP) report (No. 50-483/

82-22) prepared by the NRC Office of Inspection and Enforcement and 8311080205 831104 PDR ADOCK 05000483 O PDR

l i

issued in April, 1983.

3. In particular, the IDIP report indicates that Bechtel's civil-structural engineering group had recalculated the original design Floor Response Spectra (FRS) based upon as-built conditions and determined that in some cases the revised FRS cxceeded the original design spectra. The IDIP report further indicates that such revised TRS had not been formally transmitted to other engineering disciplines.

The Appeal Board, in its Memorandum and Order, asks what has been done since the December, 1982 I & E inspection with respect to determining if the loads imposed by the revised spectra exceed the design loads, and the safety implications, if any.

4 This Affidavit is organized into the following four sections:

Technical Explanation and Relationship to Embedded Plates; Historical Background; Current Status; and Conclusions.

Technical Explanation and Relationship to Embedded Plates

5. A floor response spectrum represents the maximum dynamic

~

response, as a function of frequency, of a single-degree-of-freedom system when excited by an acceleration time history. More simply, a floor. response spectrum represents the basic seismic load input (i.e.,

acceleration) of an item at a given location in a structure and is used for design and qualification of seismic Category I systems, com-ponents, and structures. Approximately 1200 floor response spectra i curves (which are graphical plots of the acceleration versus frequency)

I were generated for the SNUPPS Project. Although FRS curves are derived from a seismic analysis of the building structure they are neither l

l

l'

_3 dependent upon nor a function of the load-carrying capacity of embedded plates which may, in some cases, support portions of the building structure. Design loads for embedded plates are derived from a num-ber of different loading combinations, not all of which include seismic loads in the controlling combination. Since the FRS curves are used to develop seismic loas- the potential exists that revised FRS curves might result in higher loads on supports for systems and components which, if attached to embedded plates, might result in higher loads on the plates than those for which they were originally designed.

Historical Background

6. The original seismic analysis of the power block structures, performed in 1476, was based on th'e structural configuration and foundation (soil) properties established at that time for the various SNUPPS sites. As a result of this seismic analysis, FRS curves were generated and issued by the Civil engineering discipline for use in design and seismic qualification of Category I systems, components, and structures. These design FRS curves were generated for each of the.

four SNUPPS sites (Callaway, Wolf Creek, Sterling and Tyrone) in three directions (north-south, east-west, and vertical) at each floor level (masspoint) in the building structures. They were issued for use in late 1976 and early 1977 and were used by all engineering disciplines in the design of systems and procurement of components.

7. In mid-1979, it was determined by the Civil engineering dis-cipline that, based on as-built foundation (soil) properties and

i l

finalized structural configuration, as well as refinements in the

, seismic modeling of the power block, it was prudent to revise the I analytical model used for seismic analysis and evaluate the impact of any resultant changes in structural response. This revised s~elsmic I analysis was completed in late 1980 for the Callaway, Wolf Creek and 1

Sterling sites (current licensing commitment), and new as-built l FRS curves were generated and preliminarily compared with the design i

FRS curves during late 1981 and early 1982. In general, this prelim-inary comparison indicated that the as-built FRS curves were enveloped by the design FRS curves. However, the comparison also indicated that, at certain frequencies and damping values, many as-built FRS curve accelerations were higher than the design ??S curve accelera-tions. Most of these excesses were very limited in either magnitude or affected frequency range. The as-built curves were not issued for use to other engineering disciplines because it was judged that the original design curves represented conservative seismic load input. This judgment was based on the overall comparison of the i . curves, our knowledge of conservatism in the generation and appli-cation of FRS curves, experience in resolving FRS curve revisions on other projects, and review of the magnitude and range of the FRS curve excesses. Conservatisms in the design FRS curves and the as-built FRS. curves included a safe shutdown earthquate (SSE) ground acceleration level of 0.25g (versus a licensed level of 0.20g) and conservative mass distribution in the analytical models. l 1

l l

, _ _ _ _ _ _ . .- - - _ _ , , - ~ . . _ - - , _ _ _ - . _ ~ , _ . , _ - _ , ._.u . __ , .

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

i

8. As a result, the Civil engineering group began an assessment

of the excesses in mid-1982 in an attempt to establish that sufficient l conservatisms existed in the design FRS curves to preclude the need for i

issuance of revised curves. This assessment consisted of the prepara-tion of seismic load comparison calculations for the seismic Category l

I structures and the generation of detailed comparison graphs using computers and computer aided drafting. This effort was undertaken because the issuance of revised curves, when not warranted from a technical standpoint, could represent an unnecessary major impact on project cost and schedule. The assessment effort continued through early 1983, and it was during this effort that the NRC Integrated t

i Design Inspection took place (November, 1982).

9. In early 1983 it was decided that the ongoing assessment j methodology would not conclusively resolve every excess noted in the comparisons. There' fore, several alternative approaches were consid-

ered. These included

(a) Issue the as-built FRS curves for use.

(b) Generate refined as-built FRS curves by redoing the seismic analysis with struc-

tural configuration more definitively j modeled (using more masses distributed realistically in the structures) and l using an input seismic motion at the licensed SSE level of 0.20g.

(c) Establish a task force to assess the impact of the as-built FRS curves on systems, com-ponents, and structures which were designed l

1 I .

1

! (c), Continued or qualified using the design FRS curves.

This assessment would be' handled by performing an applications review (as opposed to compar-ing seismic analysis output) to establish the conservatism in original design and qualifi-Cation. ,

It was decided in mid-1983 to pursue the latter two alternatives in a tandem effort to resolve concerns regarding the excesses.

4 Current Status

10. Generation of the refined as-built FRS curves was completed in mid-October, 1983. These curves eliminated or reduced the excesses 3

I previously addressed. These refined curves will be used in the on-

] going evaluations of an FRS task force which was established in July, 1983 to review the application of the FRS curves to 1) structural components (building structures and supports other than pipe supports),

2) equipment, and 3) pipe stress analyses (including pipe supports).

The objective of this task force is to evaluate the seismic design and qualification of all seismic components of the SNUPPS plant in order to insure that they are adequate to satisfy the requirements of the as-built FRS curves.

11. The task force has completed its review of approximately ninety percent (90%) of the structural components and equipment. The i

seismic qualifications of all reviewed items meet the requirements of the as-built FRS curves.- The completed reviews encompass the loads l on all 225 safety-related manually welded embedded plates installed in the Callaway plant prior to June 7, 1977. There are no load increases on any of these plates as a result of the as-built FRS curves.

L . , ., . _- - - .

m

.~

The ma.ximum design loads on these plates are controlled by a loading combination which does not include seismic loads. In fact, the seismic design loads on these plates are reduced because the accelerations from 4

the as-built FRS curves are lower than those used in the original design for the applicable plate locations and structural frequencies.

Conclusions

12. With respect to the safety-related, manually-welded embeds installed prior to June 7, 1977 at the Callaway Plant, there has been a definitive assessment of the safety significance of the difference between the FRS used in the original design and the revised FRS. The conclusion of that assessment is that the revisec FRS curves do not result in load increases on those plates and, therefore, have no safety significance relating to those plates. Consequently, the Applicant's conclusions before the Licensing Board and the Appeal Board on the acceptability of those plates are not altered. (Refer-ence is made to Footnote 4 of the Appeal Board's Memorandum and Order.

The response provided at oral argument was based upon the original design FRS. As indicated above, however, the answer is not affected by consideration of the revised FRS for any of the 259 manually-welded

embeds, 34 of which are not safety-related.)

o .. s' h

,I ./ / e /

(/au@nt s ' }f. ._DQ' ,,JY)w2

)

Euge/ne W. Thomas Subscribed and sworn to before me n /

this J /tc1 day of November, 1983.

m -

, uo ) , ,,

A

( 'Y[tlaiTL 0 LLUDLt. - >~

Notary Public My Commission expires .

.