ML20024C583

Forwards ECCS Analysis Unit Activities Rept,Dec 1978.

ML20024C583
Person / Time
Site:	Three Mile Island
Issue date:	01/02/1979
From:	Dunn B BABCOCK & WILCOX CO.
To:	Womack E BABCOCK & WILCOX CO.
References
TASK-03, TASK-06, TASK-3, TASK-6, TASK-GB GPU-2202, NUDOCS 8307120839
Download: ML20024C583 (3)

Text

_ _ _ . ._ _

. *N ;*;, ' g (SU SX. 50,0%

"' B,ABC0%K & WILCOX COMP ANY ~

. GENERATION GROUP 3 MACK, MANAGER, PLANT DESIGN

" 8 "8 N, MANAGER, ECCS ANALYSIS (2138) Me File No.

or Ref.

Date DECEMBER ACTIVITIES REPORT JANUARY 2.1979 w.

[ m i ... .. ..... ... ...... .. ... ..w Attached please find the December Activities Report for the ECCS Analysis Unit.

BMD/lc cc: R.C. Jones L.M. Smalec J.R. Biller H.A. Bailey ECCS Analysis Unit 8307120839 790102 PDR ADOCK 05000289 C0%  : H X P HOL CO.NiS,.:_ ON. Y _

i TC5853

.--Q",

4+

g. .

ECCS ANALYSIS UNIT l

ACTIVITIES REPORT December 1978 l

'4PLISHMENTS

.I Scheduled Activities A. Loading Syndrome - A presentation on the results of Phase I was held with NRC on December 12th. The NRC seemed aware that load-ing problems were uncovered and was not concerned over this.

Informally, they have received a copy of our Phase I report.

The main topic of the meeting was on the procedures to be fol-lowed in Phase II (detailed calculations). Some specific con-cerns were raised but for ECCS the large issue was that no formal approval for the cavity pressure techniques exists. We will need to file something to obtain this approval in short order. Negotiations with licensing will go on in January and some filing occur by the first of February.

B. 205 Auxiliary Feedwater level Control - An anal performed for small breaks with a low (5steam foot)generator ysis has been level and an upgraded HPI flow. The results were quite accept-able with 13.5 feet of mixture in the core as compared to 9.5 i

but without any system changes. _

! Some system change is necessary as 9.5 feet is unacceptable.

Two changes are currently being discussed:

1) Increase level in generator to 20 to 40 feet.

2) Increase HPI flow.

These results demonstrate the viability of the second approach although cost has not yet been determined. A review of Decem-ber's fiasco on the Toledo contract, reported in unscheduled activities of this report, points out the falacy of using a I non-ECCS system to solve an ECCS problem. With these results ECCS highly recomends the upgraded HPI over the secondary level as the 205 fix. A full report on these results and ex-trapolation to other break sizes will be issued by 3/1/79.

Even if costs were considerably higher for the HPI fix we be-lieve it to be preferable.

C. Conn Yankee - Our model work is 75% complete and the key need date of 2/1/79 will be met. An approximation technique for

upper head injection has been worked out by Technical Staff l

and will be evaluated in January. In all, it would appear that a final model will not be available until mid-1979 but that suf- l ficient work has and will be perfomed by 3/1/79 to keep the l customer happy.

TO5854 L

0 3 ?-1.-

r' . ,

P inn to EA Womack Page Two December Activities Report January 2,1979 '-

D. Sn Il Break Standard Problem - The NRC has officially slipped ,

the submittal date from 1/15/79 to 4/2/79 and ECCS has made .

corresponding changes in our schedules. Model completion has been scheduled for 2/15/79. This should allow one month for parameter studies prior to filing. ,'

t E. BBR LOCA Loads - A technique for reducing BBR hot leg break ) ,

LOCA loading has been developed. It requires detailed noding inside the upper plenum of the reactor vessel. Convergence  !

studies are being perfonned at this time to decide on the exact level of detail required but an overall reduction of 40% is reasonably assured. BBR representatives will meet with ECCS on  :-

January 8th to review the work so far and to set the final sched ,,

ule ar) work content for the loads program.  !

1.2 Unscheduled Activities ,

A. Toledo Problem - Just prior to Christmas vacation, the NRC in- -

spector on site raised concerns about the acceptability of and ,

operating procedure which required operator intervention to -

i stop the auxiliary feed water filling of the generators. The , ;

NRC's main concern.was that filling the generators past 3 feet a

. would cool the primary system to the point where the pressurizer "

would drain and steam be introduced T6 the primary system. 3 Several people were involved with late night work and the end result was a bounding calculation which though very conservative . j.

showed no problems with the predicted RCS void fraction for a ~ ~ ~ -

loss of main feedwater accident with no proper operator action.*

Toledo suffered no operation restriction and the problem should ,'.'.

i l be wrapped up in January. <

, :p#

It should be noted that the cause of overfill potential was the T; -

need to have more than 3 feet level in the event of a small 0-break. This does not allow the setting of level according to ' J-a "best possible for normal operation" criteria. The use heres U .- ..*

i of a non-ECCS system to solve an ECCS concern posed the addi- t'f ,

tional problem for the non-LOCA. -

~

fixing, I believe we should not allowAlthough Toledo practices such design may be beyond to ,

g(. '

continue into other plants and we should seriously rethink our . :.

stop gap approach to ECCS design. If we wish to improve nonnal [.

operation, there should be no connection or limitation on non-

• ECCS systems. These systems should .be allowed to do the best possible job and, if necessary, ECCS modifications should be ,

made to make up for deficiencies.

• II. PROBLEMS ( .

None q III. ADMINISTRATIVE h Ken Fitzgerald and Steve Kellogg will be reporting as Co-Op's in January. '

T05855

--- - -_