ML20031H346

Testimony of Sp Congdon Re Doherty Contention 15 on Wigle Computer Code.No Basis for Contention.Criteria Contained in Spert Rept Irrelevant to Scram Reactivity Calculations

ML20031H346
Person / Time
Site:	Allens Creek File:Houston Lighting and Power Company icon.png
Issue date:	09/18/1981
From:	Congdon S GENERAL ELECTRIC CO., HOUSTON LIGHTING & POWER CO.
To:
Shared Package
ML20031H319	List: ML20031H318 ML20031H340 ML20031H346 ML20031H353 ML20031H364 ML20031H368 ML20031H379 ML20031H400 ML20031H442 ML20031H452 ML20031H458 ML20031H467 ML20031H473 ML20031H475 ML20031H479 ML20031H482 ML20031H483
References
NUDOCS 8110270390
Download: ML20031H346 (6)

Text

,

September 18, 1981 1

UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION 2

BEFORE THE ATOMIC SAFETY AND LICENSING BOARD 3

In the Matter of S

4 S

HOUSTON LIGHTING & POWER COMPANY S

Docket No. 50-466 5

S (Allens Creek Nuclear Generating S

6 Station, Unit 1)

S 7

DIRECT TESTIMONY OF STEVE P.

CONGDON 8

REGARDING'DOHERTY CONTENTION NO. 15 - WIGLE CODE 9

Q.

Would you please state your name, and your position, 10 and describe your educational and professional background?

11 A.

My name is Steve P.. Congdon.

I am employed at l~'

General Electric Company as a Nuclear engineer.

My educational 13 and professional background is described in Attachment SPC-1.

Q.

Doherty Contention No. 15 alleges that the computer 14 15 code used by the General Electric Company to predict SCRAM 16 reactivity following a Power Excursion Accident (PEA) is not 17 conservati.ve, because GE's code produces results comparable 18 to the WIGLE Code.

Is i.here any basis for such a contention?

19 A.

No.

A, Holtzclaw and Dr. Williams have already testified the PEA ceferred to here is a rod drop accident.

20 This accident is not analyzed by the GE equivalant to the 21 WIGLE Code.

22 C.

Mr. Doherty cites as a basis for this contention 3

the Special Power Ex;;rsion Tests (SPERT) performed by the

~

8110270390 010918' PDR ADOCK 05000466 PDR

_1_

1

1 Idaho Nuclear Experimental Laboratories (in particular those 2

test results reported as No. IN-1370.

Do these tests show, as 3

he alleges, that the GE code is not conservative in calculat-4 ing SCRAM reactivity?

A.

Mr. Doherty apparently does not understand the 5

concept of SCRAM reactivity.

SCRAM reactivity is a measure 6

of the amount of negative reactivity produced by rapidly o

inserting the control rods, which shuts down the reactor, and is usec as an input to the analysis of abnormal transients such as turbine trip, generator loaT rejection, and main steam isolation valve closure.

General Electric uses a one-11 dimensional time / space code (ODYN) to predict the value of 12 SCRAM reactivity for various abnormal transients over core 13 life.

The code models neutronic and thermal hydraulic changes in the core which occur throughout the transient.

A one-15 dimensional model has been shown to be appropriate by detailed 6

reactor transient tests performed at Peach Bottom 2, an 1

17 operating BWR where the data from the heavily instrumented 18 core revealed the flux response to be one-dimensional.-1/

Thic 19 o de which is used to lculate SCRAM reactivity in the core 20 as a function of time following the initiation of the abnormal 21 t rar..> ie n t, was used to successfully calculate the Peach 22 Bottom reactor test data.~2/

General Electric has been very conservative in its 23 evaluation of SCRAM reactivity.

The values used fc' SCRAM 24

=

1 reactivity in calculating th'e severity of the abnormal transient are at least 20 percent less than those calculated 2

by the one-dimensional space / time code.

In addition, the 3

control rods are assumed to move at their technical specifica-4 tion speeds, whereas plant measurements have demonstrated 5

the actual performance to be much faster.

The overall conservatism employed in the transient calculations is i

demor.strated by comparisons with actual plant data generated 8

in numerous plant start-ups, as reported in ' Analytical Methods 9

of Plant Transient Evaluation for the GE BWR," NEDO-10802, 10 Vols. 1 and 2 (April, 1973).

11 Q.

Is Mr. Doherty correct in relying on IN-1370 as l~'

a basis for disputing the conservatism in GE's one-dimensional 13 time / space code?

A.

No.

The SPERT project referred to in the contention 14 15 tested the ability of the WIGLE code to calculate the time 16 behavior of a r ise of neutrons deposited in a long thin 17 multiplying assembly.

The experiment, performed in a test i'

18 reactor which bears no resemblance to a BWR core, shtwed 19 that the WIGLE code underpredicted the response to a positive insertion of reactivity.

No control rods were inserted, so 20, the test did not measure the effects of SCRAM reactivity.

21 on could argue that since it underpredicted the response to 22 p sitive reactivity insertion, it would also underpredict the 3

ne ative reactivity response caused by control rod insertion, r

tirs indicating the WIGLE cdde to 'lx2 conservative for SCRAM 1

reactivity.

However, it is my assessment that the SPERT 2

periment is so far removed from prototyoical BNR SCRAM x

3 conditions that it cannot be used for the assessment of the 4

conservatism of the WIGLE code or General Electric's one-dimensional code for SCRAM calculations.

6 In summary, although General Electric's one-7 dimensional code may in some circumstances--for the specific 8

purpose of predicting SCRAM reactivity--produce results 9

similar to results obtained from the WIGLE code, the criteria 10 contained in the SPERT report (IN-1370) are irrelevant to 11 SCRAM reactivity calculations, whether performed by WIGLE or 12 General Electric's model.

13 14 15 16

~ 17 18 19 20 21 22 23 24 f

.. ~.

References l_/

L. A. CE.rmichael and R. O. Niemi, " Transient and Stability Tests at Peach Bottom Atomic Power Station Unit No. 2 at End of Cycle 2," EPRI NP-564 (June, 1978).

2/

" Qualification of the One Dimensional Core Transient Model for BWR's, NEDO-24154, October 1978 (Vol. 2).

1

Attachment.SPC-1 l

l STEVEN P. CONGDON' l

Steven P. Congdon obtained a B.S.

in physics from Valparaiso University in I?62 and a PhD in Nuclear Engineering from Pennsylvania State University in 1966.

From 1966 to 1976 he was employea at Knolls Atomic Power Laboratory in Schenectady, New York where he developed improved methods for calculating nuclear cross sections and power distributions in Naval Reactors.

In 1976, he transferred to the Systems Dynamics Methods group at GE-San Jose where he supervised the development of a one-l dimensional nuclear-thermal hydraulic transient model for Boiling Water reactors.

This work included development of the basic equations, coding the computer model and qualification l

of the model against data obtained from tests performed at operating BWR's.

Descriptions of this work appear in four papers delivered at technical society meetings and in a number of reports submitted to the Nuclear Regulatory Commission.

t Since 1980, Dr. Congdon has held the position of-Manager, Nuclear i

Methods and has the responsibility for GE's steady state nuclear design technology for BWR's.

l l

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