ML20038B591

Submits Info Requested in 811106 Telcon W/Nrc Re Turbine Missile Shield

ML20038B591
Person / Time
Site:	Atlantic Nuclear Power Plant
Issue date:	11/12/1981
From:	Haga P OFFSHORE POWER SYSTEMS (SUBS. OF WESTINGHOUSE ELECTRI
To:	Adensam E Office of Nuclear Reactor Regulation
References
FNP-PST-095, FNP-PST-95, NUDOCS 8112080418
Download: ML20038B591 (4)

Text

O -

Offshore Power Systems uma w , 1 E x, ,

w w;m noo Mc t P'O ! > :n ' .E!16 % U211 T+! - ' M4C6 .

6 h

November 12, 1981 V o,g<r/g x\ -

e Ms. Elinor Adensam, Chief -

s % v

• j *-.

Licensing Branch No. 4 U.S. Nuclear Regulatory Commission 3 k 6

'2d 7920 Norfolk Avenue / TQq% 2)

Bethesda, MD 20852 s\

f<f!ilmi\ 6 Re: Docket No. STN 50-437; P. B. Haga Turbine Missile Shield Evaluation Director

% .rsy, tom ,T a n =4 *

Dear Ms. Adensam:

This letter provides the subject information requested by the staff in a telephone conversation on November 6, 1981.

Tests conducted by EPRI in 1978 showed that turbine missile energy loss during penetration of steel barriers might be less than predicted by at least one empirical fonnula (BRL). As a result of these tests, the staff reported a possible error in its testimony to the Hearing Board, namely, that the BRL formula (used by the staff) may not adequately predict missile behavior for the piercing orientation. This lead the go l staff to the conclusion that the FNP turbine missile shield may not be adequate to stop design overspeed missiles as had been stated in their testimony.

V y/

1 The 4-1/2 inch steel missile shield is included in the FNP to reduce the potential for excessive flooding of the platform from turbine missile damage. The require-ment to prevent excessive flooding is satisfied for all instances when flooding occurs in two adjacent compart-ments, because the main deck (100 foot elevation) remains above water, the platform does not ground and equipment required for safe shutdown remains operable (PDR Sections 3.12.8.2 and 15.6.2.1) . These criteria are also met for certain instances of more than two 8112080418 811112 PDR ADOCK 05000437 A PDR

~.

Page #2 - Ms. Elinor Adensam November 12, 1981 compartment flooding. Platform hull compartmentation is described in POR Section-3.12.8 and is shown in Figure 3.12-1, Sheets 1 and 2.

Whether the turbine missile shield, as well as other barriers, will stop an impacting missile depends critically on the missile impact energy and missile orientation. The staff has asked that OPS re-evaluate design overspeed missiles below the turbine center plane under the assumption of a piercing orientation for all barrier interactions. The staff has also asked that the effect of such missiles on safety be evaluated on a purely deterministic basis. OPS believes these staff requests to be unnecessarily restrictive because (a) a series of missile impacts all with a piercing crientation is very unlikely and ~ in some cases not possible and (b) the criterion for turbine missile protection has always been probabilistic in nature and has been adopted by the staff itself (PDR Section 3.5.1 and Regulatory Guide 1.115). Nevertheless, OPS has evaluated the effects of design overspeed missiles with downward trajectories under the assumption of piercing orientation where such impact is judged to be possible and has concluded as follows:

a. Failure of disc 4 or 5 at design overspeed may result in a missile which penetrates the missile shield and could penetrate the platform bottom.

b. Failure of disc 2 or-3 at design overspeed may also result in a missile which penetrates the missile shield; however, the residual energy of such a i missile would not be sufficient to penetrate the platfonn bottom shell.

c. On a purely deterministic basis, it is possible for a missile from disc 4 or 5 to cause flooding of i more than two compartm+..its.

d. Design options are available which can limit the extent of flooding to no more than two compart-ments. These options include:

i e Selecting a BB281 L. P. turbine design which increases missile energy absorbtion. A further option exists to utilize a different building block (BB) which, for example, incorporates a thicker blade ring to decrease missile exit energy.

t'.

Page #3 - Ms. Elinor Adensam November 12, 1981 o Increasing the thickness of the turbine missile shield in selected regions, for example, opposite discs-4 and 5. It is also possible to add barriers locally in the near vicinity of the turbine, for ex~ ample, adjacent to the condenser shell.

e Making existing non-watertight bulkheads watertight so as to reduce the spread of flooding. Since existing bulkheads are of welded construction, it is only necessary to . provide watertight penetrations in order to make a bulkhead watertight. Additional watertight bulkheads can al so be provided if and where necessary.

OPS does not propose to implement the design changes cited above at this time, because we believe that the present issue will be satisfactorily resolved as outlined below during final plant design. Neverthe-less, OPS commits to make the appropriate design changes if such changes are ul timately held to be necessary during the final design review. We believe that such design changes will not be necessary' for reasons which include the following:

a. Turbine design improvements presently being studied -

will reduce the probability of missile formation,

b. Calculations for existing Building Block 281 rotors show that the probabili ty of tgrbige missile Disc 1 fonnation is dominated by disc 2 - .

rupture does not result in an external missile and the probability of disc 3, 4 or 5 rupture is calculated to be decades lower than disc 2. The estimated probability of disc 2 rupture is- in reasonable agreement with the Bush estimate.

Missiles from discs 2 and 3 do not penetrate the missile shield with significant residual energy.

/I-- Westinghouse Steam Turbine Generator Division, WSTG-1-P, Procedures for Estimating the Probability of Steam Turbine Disc Rupture from Stress Corrosion Cracking (May 1981).

/2 Westinghouse Steam Turbine Generator Division, WSTG-2-P, Missile Energy Analysis Methods for Nuclear Steam Turbines (May 1681).

Page #4 - Ms. Elinor- Adensam November 12, 1981

c. The assumptions presently required by the staff for calculating barrier penetration are based on an ultra-conservative 1978 EPRI test. In this test, a missile was placed in the piercing (or arrowhead) configuration end. fired without rotation into the target. The test was intended to be (and certainly was) a bounding case. In fact, the test was un-realistic because disc rotation and blade crushing were not modeled. With' these factors taken into account, occurrence of the piercing orientation as modeled by EPRI is very unlikely, although a spectrum of possible impact areas does exist. EPRI has performed additional turbine missile tests since 1978. Some of the tests have been spin tests of discs , including blades. . Preli::linary results from these later tests _ were reported at the 1981 EPRI Seminar in New Orleans and the indications are that spinning missiles with blades are not as penetrating as the 1978 results suggested.

We trust that the foregoing will enable the staff to respond positively to the concern expressed in Mr. S.M.

Schinki's letter to the Hearing Board dated December 22, 1978.

Very truly yours,

./

P.B. Haga Director, Power Systems Technology cc: A.R. Collier V.W. Campbell