ML20081L350

Responds to 831021 Telephone Discussion Re Seismic Calculations & Related Concerns

ML20081L350
Person / Time
Site:	Seabrook
Issue date:	10/25/1983
From:	Markowitz G AFFILIATION NOT ASSIGNED
To:	Terao D NRC
Shared Package
ML20081L345	List: ML20081L350
References
NUDOCS 8311160168
Download: ML20081L350 (4)

Text

.

Ot *

• O DUKE - 5179

. CATAWBA - 2355 Westinghouse Water Reactor NUCbf CommefClal

• '*HS5 *5ba Electric Corporation Divisl6ns B0:355 PittsburthFennsylvania15230 MPS #35924 hPS #4100 July ll, 1980 Ref: DUKE-5128/ CATAWBA-2311

- Mr. W. O. Parker Vice President Steam Production Department Duke Power Company P.O. Box 33189 -

Charlotte, North Carolina 28242 Attention: G. A. Copp/R. O. Sharpe MCGUIRE/ CATAWBA h11 CLEAR STATIONS

• UNITS NtNBER 1 AND 2 Reportable Item - Centrifugal Charging Pumps

Dear Mr. Parker:

~.

The referenced letter provided information on the Part 21 reportable item on Centrifugal Charg~ing Pump operation following secondary side high energy line rupture. 'Ihe. attachments to this letter supplement the referenced infomation and provide a generic evaluation of the sensitivity of the FSAR transient analyses to the emergency operating procedure interim modifications which were proposed. De basic conclusion of the attached infomation is that the interim modifications have negligable impact on the FSAR analyses.

If you have any questions on this information, please contact G. L. Fidler.

Yours truly, bOSSSLW R. S. Howard, Manager

\W Duke Power Projects Icn cc: D. L. Fuller 1L W. O. Parker 10L,10A

! S. K. Blackley IL, 2A l_.

8311160168 831108 1 PDR ADOCK 05000443 l A PDR

.. .. ._ Attachment 1 ,

u- -

CENTRIFUGAL CHARGING PUMP OPERATION FOLLOWING' SECONDARY SIDE HIGH ENERGY LINE RUPTURE Reference 1: NS-TMA-2245, S/8/80 h Rdference 1 notified the NRC of a concern for consequential damage of one or more centrifugal cnarging pumps (CCP) folicwing a secondary system

, high energy line rupture. Reference 1 included a calculational method and sample calculation to permit evaluation of this concern on a plant specific basis. Should a plant specific problem be identified, Westinghouse provided several recommendat' ions for the interim until necessary design modifications can be implemented to resolve the problem. These recommenda-tions included two proposed interim modifications which included:

1. Remove the safety injection initiation automatic closure signal from the CCP miniflow isolation valves.

2. Modify plant emergency operating procedures to instruct the operator to:

v

a. Close the CCP miniflow isolation val.ves when the. actual RCS pressure drops to the calculated pressure' for manual reactor

! coolant pump trip.

b. Reopen the CCP miniflow isolation valves should the wide range RCS pressure subsequently rise to greater than 2000 psig.

Prior to making this recommendation, Westinghouse evaluated the impact of Y the recommended operating procedure modifications on the results of the various accidents which initiate safety injection and are sensitive to CCP flow delivery. The accidents evaluated in detail include' secondary system ruptures and the spectrum of small loss of coolant accidents. The analytical results for steam generator tube rupture and large loss of coolant accident are not sensitive to a reduction in CCP flow of the magnitude that results from the recommended modifications. This letter functions to supplement Reference 1 and identify the sensitivity of the accident analyses to the recomended modifications. This evaluation is generic in nature.

A Attachment i l]*

eaa.

( 2 O  %

Secondary System Rupture b

Sensitivity analyses have been performed for secondary high energy line ruptures to evaluate the impact of reduced safety injection flow due to normally open miniflow isolation valves. These analyses indicate an 4 insignificant effect on the plant transient response.

l

~~

A. Feedline Rupture Following a feedline rup.ture, the reactor coolant pressure will reach the pressurizer safety valve setpoint within approximately 100 seconds i

assuming maximum safeguards with the power-operated relief valves inoperable. With minimum safeguards, the reactor coolant pressure will not reach the pressurizer safety valve setpoint until'approximately 300 seconds; The time that the reactor coolant system pressure remains at the pressurizer safety valve setpoint is a function of the auxiliary feedwater flow injected into the non-faulted steam generators and the

. time at which the operator is assumed to take action. With the mini-flow isolation valves open, the peak reactor coolant system pressure and the water discharged via the pressur'izer safety valves are insignifi-cantly changed from the FSAR results.

B. Steamline Rupture The effects of maintaining the miniflow isolation valves in a normally open position was also investigated following a main steamline rupture.

h For.the condition II " credible" steamline rupture, the results of the transient with the miniflow valves open showed that the licensing criterion (no return to criticality after reac' tor trip) continues to be met. The condition III and IV main steamline ruptures were also h reanalyzed assuming the miniflow valves were open. The results of the analysis showed that, even with reduced safety injection flow into the core, no DNB occurred for any rupture.

O

, a'

\< ('  :-, L Attachment 1 A Small Loss of Coolant Accidents

\8 Sensitivity analyses have been performed to evaluate the impact of reduced safety injection flow on small break loss of coolant accidents (LOCAs).

h These analyses indicated that miniflow isolation can be delayed, but it must occur at some time into the small break LOCA transient in order to limit the peak clad temperature (PCT) penalty.

h The proposed modification delays miniflow isolation and reduces SI flow delivered by approximately 45 ijpm at 1250 psia during the delay time period. j l The impact of this modification was evaluated based on two isolation times: l

1) The time equivalent to the RCP trip time, and 2) approximately 10 minutes in the transient, or just prior to system drain to the break for the worst ,

j small break sizes. Tne second time was evaluated to determine the impact if the operator does not isolate miniflow within the proposed prescribed '

time. The spectrum of small break sizes are considered to encompass all  !

i

, possible small break scenarios. Only cold leg break locations are considered f

l h since they will continue to be limiting in terms of PCT.

i f I A. Very small breaks that do not drain the RCS or uncover the core, and '

maintain RCS pressure above secondary pressure (< s2" diameter).

For these break sizes, it is quite possible that the operator may never isolate the miniflow line, since the pressure setpoint will not be reached, and continued pumped SI degradation will persist.

liowever, this will have no adverse consequences in terms of core uncovery and PCT. No core uncovery will be expected for the degraded SI case, similarly to the base comparison case with full SI. The only effect would be a slightly lower equilibration pressure for a given break size.

B. Small breaks that drain the RCS and result in the maximum cladding temperatures (2" < diameter < 6").

O .

This range of break sizes represents the worst small break size for .

-_ _ _ _ - _ _ _ _ _ - - - - - - - _ - - - - - - - - - - - - - - - - - I

, e..

O V Attachment 1 most plants as determined utilizing the currently approved October 1975 Evaluation Model version, as shown in WCAP-8970-P-A. If miniflow is isolated at the RCP trip setpoint rather.than the "S" signal, a reduc-tion in safety injection flow of less than 45 gpm results, averaged h for the approximately 50 second period of time separating the two events.

This reduction in RCS liquid inventory results in core uncovery less than one second earlier, and has a negligible impact on PCT. If mini-flow is isolated at the time of core uncovery, or approximately 10 h minutes for b'reak sizes in this range, a greater reduction in RCS liquid inventory results in a c' ore uncovery 10 seconds earlier in the transients resulting in less than a 10*F PCT penalty for the worst size small break.

This would not result in any present FSAR small break analysis becoming more limiting than the correspc.iding large break LOCA FSAR analysis. ,

l If miniflow isolation does not occur at any time into the transient for l this category of small LOCA, a PCT penalty of 200*F or more could occur.

1

'C. Small break sizes larger than the worst break through the intermediate break sizes (1 6" diameter).

Ereak sizes in this range have been determined to be non-limiting for small break utilizing the currently approved October 1975 Evaluation Model, WCAP-8970-P-A. If miniflow isolation occurs at the RCP trip time for these break sizes, the negligible effect on PCT presented above also applies. Similarly, if isolation occurs prior to core uncovery, the small (< 10*F) PCT penalty will result as well. However, for'these larger break sizes, the time of first core uncovery occurs prior to 10 minutes. If miniflow isolation is not performed until 10 minutes, reduced SI will be delivered during the core uncovery time, which can have a greater impact on PCT. Studies indicate a potential PCT penalty of 40*F resulting for these non-limiting break sizes if miniflow is not isolated until 10 minutes. This is not expected to shift the worst break size to larger breaks, since these breaks are typically hundreds of degrees less than smaller limiting small breaks O analyzed with the currently approved Evaluation Model.

.m. ( 7 ( Attachment I  !

, 'f 5 e h For all FSAR small LOCA analyses, one complete train failure is assumed.

is clear that two charging pumps without miniflow isolation provides more It flow than one pump with miniflow isolation. The impact presented in this evaluation maintains the one train failure and assumes no miniflow isola-h tion for the remaining pump. If both pumps were operating, the PCT results would be much lower than present FSAR calculations even if miniflow isola- 1 tion is not assumed to occur for the two pump case. In this situation, the ,

plant FSAR small break calculations remain conservative.

These sensitivity studies form the basis for the recommended interim modifications to the emergency operating procedures. The accidents evalu-

~

ated are relatively insensitive to the recorrnended modifications. Further, the accidents evaluated will give results that satisfy acceptance criteria as long as the CCP miniflow is isolated within 10 minutes of event initiation.

However, small LOCA sensitivity studies with one SI train operating confirm that small LOCA analyses require miniflow isolation within 10 minutes.

To comply with the recommended modifications, the operator can isolate mini-flow at any point in the depressurization transient prior to RCS pressure reaching the RCP trip setpoint. Should a repressurization transient occur, the operator can open CCP miniflow at any point between the RCP trip set-4~* point and 2000 psig. Such operator actions will ensure that plant accidents satisfy acceptance criteria and protect the CCPs from consequential damage during the repressurization transient that accompanies a secondary system -

high energy line rupture at high initial power levels.

(%-

C .

4

Attachment 2 .-

,g

., (. . 4

, ,- CENTRIFUGAL CHARGING PUMP OPERATION FOLLOWING SECONDARY SIDE HIGH ENERGY LINE BREAK (UHI PLANT SUPPLEMENT)

The small loss of coolant. accident (LOCA) section of the main report was generated primarily for plant applications which do not include upper head injection (UHI) as part of the ECCS design. This supplement pro-vides additional small LOCA info'rmation for UHI plants and, together with the main report, assesses the impact of delayed miniflow isolation for small LOCAs for UHI plants.

T he model utilized to determine the SI sensitivities and to identify the worst small break size discussed in the main report was the October 1975 Model (WCAP-8970-P-A) version of the Evaluation Model. This model is not yet approved for UHI plant analyses. UHI small break analy'ses are performed with the December 1974 small break version. However, sensi-tivity studies performed to determine the effect of pumped SI on small break LOCA PCTs utilizing the December model yielded nearly identical results as presented in the main report. This is expected since the model changes included in the October model do not affect the basic vessel inventory and core boiloff relationships that determine the impact of changes in pumped safety injection on' PCT.

A n important difference in UHI plant small break analysis results as compared to similar non-UHI plant analysis results is the small break size resulting in the highest PCT. This break size is generally greater g for UHI plants than for similar non-UHI plants because of the additional safety injection flow'provided by the UHI accumulator at relatively high RCS pressures. The worst small break size for UHI plants may be a six inch diameter break or larger. The main report identified breaks of this size and larger as non-limiting small break sizes. While this is true for non-UHI plants, it is not accurate for typical UHI plant small break analyses. Therefore, the stated 40*F potential penalty for A)

Q

r

- ~"

..g (

U '

Attachment 2 O 4

() six inch breaks applies to the worst break for UHI plants for the case where miniflow isolation is delayed until 10 minutes. It is Westinghouse's opinion, however, that the stated penalty of 40*F is conservatively high and bounding for UHI plants, for the following h reasons: a) The 40'F penalty was based on sensitivity studies performed assuming an approximate 20% reduction in total HPI flow. However, the anticipated 20% reduction actually applies only to the charging pumps.

Intermediate head SI pumps are not affected. Therefore, total HPI for plants with intermediate head SI pumps, which includes all UHI plants, will result in less total' degradation, and thus a smaller PCT penalty.

The high pressure accumulator on UHI plants has a similar effect of reducing the total HPI degradation due to the delay in miniflow isolation.

b) The UHI accumulator is a significant source of liqujd mass inventory for breaks greater than or equal to six inches in diameter. This addi-tional mass delays the core uncovery time as compared to the same size break occuring on a~ similar non-UHI plant, since more liquid mass must exit from the break prior to core uncovery. The delay in core uncovery

)

results in clad heatup at a lower power level caused by the decay in residual core heat. Therefore, clad heatup rates are slower which also tends to reduce the sensitivity to changes in HPI delivery ratt..

In conclusion, the sensitivity provided for six inch diameter and larger break sizes in the main report represents the worst break size range for UHI plants. The stated 40'F PCT penalty for breaks of this size resultant from s 10 minute delay in miniflow isolation is a conservatively high and boimding value for UHI plants, for the reasons stated above.

If miniflow is isolated at the time of RCP trip, the negligible impact on PCT discussed in the main report applies for UHI plants as well.

The <10 F penalty resultant if miniflow isolation occurs prior to core

@ uncovery also applies to UHI plants, with the added benefit that this

• event occurs later in a UHI plant transient than for a non-UHI plant transient of the same break siz'e, allowing more time for the operator to act.

O -

-