ML19318D215

Forwards Resolutions to long-term Generic Issues Re May 1979 Commission Orders & Addl Info Requested Concerning Small Break LOCA

ML19318D215
Person / Time
Site:	Crane
Issue date:	06/30/1980
From:	Herbein J METROPOLITAN EDISON CO.
To:	Harold Denton Office of Nuclear Reactor Regulation
References
TLL-285, NUDOCS 8007080074
Download: ML19318D215 (100)

Text

'

Metropolitan Edison Company Post Office Box 48C A

Middletown, Pennsylvania 17057 717 944 4041 writer's Dicet De t Numt:er a

June 30, 1980 TLL 285 Office of Nuclear Reactor Regulation Attn:

Harold R. Denton, Director U. S. Nuclear Regulatory Commission Washington, D.C.

20555

Dear Sir:

Three Mile Island Nuclear Station, Unit I (TMI-1)

Operating License Number DPR-50 Docket Number 50-289 Resolution of Long-Term Generic Issues Related to the Commission Orders of May 1979 The items related to the long-term porti<n of the Commission Orders generic to all B&W plants, as listed in Enclosure 1 of Reference 1, and additional infor-mation requested concerning small break LOCA, Reference 2, have been reviewed for the Three Mile Island Unit I.

• 1i' resolution for each of the items is listed in the attachment to this let Most of these issues have been oddrecsed in t.e TMI-I Restart Report.

Reference to the pertinent sections of the report is giv,n for the items.

In some cases, letters have been sent to the NRC resolving these issues.

For those items, the reference letter number and dates for the letters have been identified.

inc re y, 0

J. G. Herbein Vice Dresident TMI-I JGH:MI:1ma O/

Enclosures 3

///

sn udosae k-p(qd,ay h h Wa afla C/~.Q THIS DOCUMENT CONTAINS POOR QUAllTY PAGES 8007 08007q

\\te'rcCchtan Ed SCn CcmCany :S 3 '.'eTCdf Of :re Gerera; Reac Uttt es 5,s:em

1. R. Denten TLL 285 cc:

B. J. Snyder J. T. Collins D. Dilanni H.-Silver B. H. Grier REF:

1.

NRC letter from D. F. Ross, Jr. to all B&W Operating Plaats except

+

3 TMI-I and TMI-II, " Identification and Resolution of Long-Term Generic Issues Related to the Commission Orders of May 1979,"

dated August 21, 1979.

2.

NRC letter from R. W. Re to all B&W Operating Plants except TMI-I and TMI-II, " Request for Additional Information - Small Break Loss-of-Coolant Accident," dated November 21, 1979.

t 4

t,

I A

.+y-r

ATTACIIMENT 1

i Items related to the long-term portion of Commission Orders generic to all B& W operating plants.

(Enclosare 1 of Reference 1).

t 0

1 Item'No. 1:

(Reference 1) f Failure mode and affects analysis of the integrated control system.

B&W has indicated that this repdre will be available for our review by August-20, 1979. By August 31, 1979, each licensee should endorse this report, or indicate the degree to which it is not applicable.

Following our staff review of this report, any system or procedural changes necessary will be sett to each licensee.

Respong :

The B&W's Generic Report BAW-1564, " Integrated Control System Re-liability Analysis," has been endorsed by Met-Ed/GPU by Letter No.

E&L-1896 dated October 26, 1979. Also, this item is discussed in the TMI-1 Restart Report, Supplement 1, Part 3, Response to Ques-tion 12.

f 6

a 4 -:.y;;+~~gryL

=

-..-~a---.c

-. :r,-

c,nnm u.

m

- _It'em No. - 2: - (Reference No. 1)

-Continued operator training and: drilling.

i Each licensee shall document the steps'it has taken to insure that con-tinued operator training and drilling incorporates the necessary lessons learned from TMI-2 and assures a continuing high state of preparedness.

This shall be submitted to the NRC by September 27, 1979. Pending Com-mission action regarding improvements in the Operator Licensing Program, this requirement may be keyed to an upgrade in the initial training and requalification program by licensees.

Response

The above issue has been discussed in detail in Section 6 of the TMI-1 Restart Report and in the response to Question 49 of Supplement 1, Part 1 of the Restart Report.

e i

I I

i 4 '

dr4 gq.

1O a-e@-

a%

w ar-w a-

Item No. 3:

(Reference.1)'

Upgrade of.the anticipatory readtor trip to safety-grade.

i Each licennee has submitted a preliminary design for implementing a safety-grace reactor trip upon loss of main feedwater and/or turbine trip.

The staff is evaluating these proposals at.the present time.

Staff comments will be issued to each licensee by August 31, 1979.

In light of the'recent failure of the control-grade trip at ANO-1, accelerated installation schedules should be developed.

Response

Reactor trip upon loss of main feedwater/ turbine trip has been discussed in Section 2.1.1.1 of the TMI-l Restart Report and is i

a safety-grade installation.

4 f

I-i i

i

.a f

4 9

r

, s 1

.-c y

i Item No. 4:

(Reference 1)

Auxiliary / emergency feedwater system reliability upgrade.

.i The long-term provisions of-the Orders vary on this requirement. We believe that the most efficient way to fully define the needed improve-ments is to perform the AFW/EFW system reliability study discussed in our July 19 and August 9, 1979 meetings with the Owner's Group.

By August 17, 1979. we expect a letter from B&W outlining in detail the scope of the study and the schedule for completing the study.

By Au-gust 31, 1979, each licensee will submit a letter to the NRC committing to the proposed schedule and study, or provide an alternative.

The study for the lead plant (tentatively Rancho Seco) will be available for our review in draft form by September 17, 1979. The studies for the remaining plants will be available in draf t form by October 22, 1979.

The final report will be published by December 3,1979.

Response

The Auxiliary / Emergency Feedwater System upgrading has been discussed in Section 2.1.2.6 and in response to Question 4 in-Supplement 1, Part 3 of the TMI-l Restart Report.

In addition, the EFW reliabilicy study generic and TMI-l specific was forwarded rn the NRC by letter (E&L-2102) dated February 7, 1980.

I 4

4 1

4

~

I i

~

Item No. 5:

(Reference 1)

A detailed analysis of the thermal-mechanical conditions in the reactor vessel during recovery from small breaks with extended loss of all feed-water.

This issue was identified in the staff evaluations for Rancho Seco, Davis-Besse 1, and Crystal River 3.

However, it is also applicable to Oconce and Arkansas Nuclear One 1.

Our request for additional informa-tion on this subject was sent to Mr. J. H. Taylor (BaW) from Mr. D. E.

Ross (NRC) by letter dated July 12.

In a letter from Taylor to Ross dated August 3, 1979, B&W stated:

" Prior to responding to your letter (dated July 12), we feel it is essential to have discussions with our utility customers. Following this discussion, we will provide you with a schedule." We desire this schedule from the B&W utilities by August 31, 1979.

Note:

It appears to us that the concern is valid for Davis-Besse, but to a lesser degree due to the significantly lower shut-off head of the HP1 pucps.

Response

This issue is the shbject of current discussion with NRC and B&W Owner's group. This is still an open item and is b2ing investigated as a generic issue.

Item No. 6:

(Reference 1)

PORV and safety valve lift frequency and mechanical reliability.

i This item is discussed in Section 8.4.6 of NUREG-0560 and endorsed in the staff's evaluation for each plant. This requirement has been super-seded in scope and schedule by Recommendation 2.1.2 of NUREG-0578.

Li-censees will be directed by letter to take further action on this matter in the near future.

Response

The TMI-l PORV actuation frequency has been discussed in the following letters to the NRC:

A.

Letter CQL-1439 dated November 27, 1979, and B.

Letter TLL-028 dated January 18, 1980.

Item No. 7:

(Reference 1)

Small Break LOCA Analysis.

i This item is discussed in Section 8.4.2 of NUREG-0560 and endorscd in the staff's evaluation for each planc. Most of this work has been com-pleted for the B&W plants.

Ilowever, additional information is still required before the staff can issue its evaluation (NUREG-0565

" Staff Report on Generic Evaluation of Small Break Loss-of-Coolant Accident Behsvior for Babcock & Wilcox Operating Plants"). Attachment A to this enclosure is a listing of the specific information needed. We plan on issuing NUREG-0565 in late September 1979.

By August 31, 1979, provide a schedule for the submission of Items 1 through 5 of Attachment A such that the information will be received in time to support the publication of NUREG-0565.

R2sponse:

Much of the information required to respond to this question was gene-rated by B&W for the B&W Owner's Group. Generic and plant specific response to Items lA through 6 of Item No. 7 are provided in the fol-lowing pages.

t L

Item No. 7: Attachment A to Enclosure 1 of Reference 1 1A.

Provide a benchmark analysis of sequential auxiliary feedwater flow to the steam generators following a loss of main feedwater. This analysis was provided in a letter from J. Taylor (B&W) to R. Mattson (NRC) dated June 15, 1979. Ilowever, in this analysis, the TRAP-2 code with a 6 node steam generator model was utilized. All small break analyses presented to the NRC have been performed using the CRAFT-2 code with a 3 node steam generator model. We require a benchmark analysis for sequential auxiliary feedwater flow also be performed using CRAFT-2 with a 3 node steam generator represen-tation.

Response

See Attached 1

1 n

A i

The Babcock & Wilcox Company Nucicar Power Generation Division s

1 0

6 6

s,

C0tlTEilTS "IflTRODUCTI0:1 II SITE EVEtlT DESCRIPTIO!! 3

[-

III METi!0DS IV-RESULTS V

C0.CLUSI0tlS s

FIGURE I CRAFT-2 fl0DIt'G DIAGRAM FOR SMALL BREfKS FIGURE 2 STARTUP FEEDWATER FLOW

~

FIGURE 3 STEA!! GENEP TOR LIQUID LEVEL (TEMPERATURE ADJUSTED)

FIGURE 4

- STEAM GENERATOR SECONDARY SIDE PRESSURE FIGURE 5 PRIMARY A LOOP TEMPERATURE FIGURE G PRItRRY B LOOP TEf!PEPATURE FIGURE 7 PRESSURIZER LEVEL FIGURE 8 REACTOR VESSEL PRESSURE

~

e O

6 e

8

+

e e

4 0

1-4-80

~~--

m._..

3-

~,

~'

I.

INTRODUCTION This report

• presents an analysis of sequential auxiliary feedwater (AFW) flow to the once-through steam generators for a loss of main feedwater transient. The CPAFT2 code,I and the small break model; described in reference 2 have been used in the study. The calculated results have been compared to a loss of offsite power startup test data obtained from the Florida Power Corporation's Crystal River 3 Unit in.which an imbalance in the auxiliary feedwater flows between the two operating loops resulted in an imbalance in the primary loop response.

This transient tests several features of the computer simulation, including conditions of asymmetric loop temperatures, an almost dry generator to feed auxiliary feed-water into, loss of RC pumps, and establishment o.f natural circulation. In many i

cases the absolute validity of the boundary conditions and test data were ques-tionable, and cetimates had to be used.

Ilowever, - this analysis does' show that the data trends can be predicted by a 3 node CRAFT 2 SG representation.

II.

SITE EVENT DESCRIPTION The Crystal River 3 Unit is a 2452 MNt, 177-FA B&W reactor.with a lowered-loop configuration. On April 23, 1977, a loss of offsite power test was performed.

This test was initiated from approximately 15% full power operation.

The secon-

. dary liquid levels were approximately 2 feet and was sufficient to remove the powet and provide essentially steady-state operation prior to test initiation.

The test was initiated by tripping the reactor, the reactor c'oolant pump, and feedwater pump power sources.

The core power then dropped to the decay heat level and, as the primary coolant pumps coasted down, the primary flow decayed to natural c'irculation 1cvel.

One diesel generator.was started to provide power for the pressurizer heaters, one makeup punp, and other necessary services of secon-dary importance to this analysis.

The main feedwater flow coasted down resulting in both steam generators eventually drying out until the auxiliary feedwater flow became sufficient to start filling the A loop steam generator secondary at about two minutes into the transient.

The B loop steam generator remained dryed out until twelve to fourteen minutes into the transient when the A loop reached. normal operating level and the feed-l water flow was diverted to the B loop. The imbalance in the feedwater flows, and hence 1cvels, resulted in a corresponding imbalance in the primary system re-sponse including the decay heat removal, the hot and cold Icg temperatures and

- '. N12--

i.

I

~

6 e-

--n-,

r n-,-,-

..->s

flows betucen the.two loops.. The transient results were used to evaluate the ability of the 3 node CIMFT2 steam generator model used 'in small break evalua-tiens to calculate the effect of the feedwater transient.

III. '!IETHODS

~

A.

CRAFT Input Model The input model developed for this calculation was based on the small break model

.used for licensing.2 The schematic of the flow path nodalization is shown in Figure 1.

The initial system conditions were defined based on the available mea-cured data which vere required to represent this test.

The nodel was set up to provide a steady-state calculation until tuo seconds into the transient when the reactor, reactor coolan; pumps, and main feedwater pumps vere tripped initiating the transient calculation.

B.

Initial Conditions The initial mass flow was assumed to be identical to the full power operation a

value.

The measured hot and cold leg temperatures were then used to determine a consistent core power to provide the initial steady-state operating conditions.

This resulted in an initial power of 19% of full power operation versus the 15%

power defined in the summary test report. 'lland calculations, using the 15% core power,and the measured hot and cold temperatures, resulted in a mass flou con-siderably below that required to balance the pump power.

The, actual mass flow is believed to have been only 1 or 2% less than full power flow.

The pressure dis-tribution around the system was revised, because of the new hot and cold leg temperatures, to maintain the loss coefficients defined by the referenced model.

The liquid levels in the pressurizer and steam generator secondary were changed to reficct the measured data.

C.

Boundary Conditions The makeup pump flow was modeled by defining the pressure flow characteristic curve for normal operation with the recirculation line open.

The makeup pump wan actuated when the pressurizer level dropped to 30" below the initial liquid level value.

The makeup pump flow was equally distributed between the two cold leg pump discharge modes as shown in Figure 1.

The feedwater flogs were defined by the test data and are given in Figure 2.

An auxiliary feedwater enthalpy of 58 btu /lbm, which is the nominal enthalpy of the system,was used.

"Thsi safhtyreliefvalvesw2resetto1030 psia.tomodel~thecffectoftheturbine bypass valvas, which are fully open~at 1030 psia. The safety relief flow is the only_ allowance made in the model for steam flow.

Quie heat transfer to the secondary was assumed to be ~ to the mixture in the lower portion of tiie cteam generator and the! fraction which may have been deposited in

~

the steam region was assumed to bc negligible. A preliminary sho'rt-term transient evaluatio'n demonstrated the need to define the heat transfer multiplier based on the-steam generator secondary levels.

Consequently, the final model contained a heat transfer multiplier as a function of time based on the measured secondary levels.

.IV.

RESULTS This section presents-a comparison of the CRAFT 2 analysis to the data taken for the first 20 minutes of the CR-3 loss of offsite po"er test.

As will be shown, some of the data utilized in the evaluation is questionable and greatly ir fluence

~

the transient response.

Ilowever, even with the unce-tainties in the measured data, the CRAFT 2 code is shown to adequately calculate the RCS behavior.

A.

Sc'indary Response Figure 3 shows the secondary side SG Icvels'during the test.

The test data shows that, following the loss of main feedwater, the initial level in both steam gen-erato'rs decreases. At approximately 1' minute into the transient, the auxiliary feedwater system initiates, as shown in Figure 2, and preferentially feeds the A loop steam generator.

Thus, the liquid 1cvel in SG A increases.

At 12 minutes, the liquid level in SG A stabilizes because it has reached its control point. At that time, the feedwater flow ~is diverted to SC B and its level increases.

The CRAFT 2 c' ode calculated results shows reasonable agreement with the SG A level during the first 12 minutes.

After this time, however, the CRAFT 2 calculation continues'to increase the SG 1evel while the data shows a level stabilization-after'this time.

This difference is probably due to an overestination of the auxiliary feedwater flow to SG A after this time.

The auxiliary feedwater flow, as indicated in Figure 2,'is vet, stable and at a relatively high flowrate af ter 12 minutes.

Examining other-data, such as the A loop hot and cold leg tempera-tures, does not support achigh auxiliary feedwater flowrate.

In light of the

. ability-of the CRAFT 2 code to reasonably predict the SG response up to 12 min-utes and the inferences obtained from other data, the flowrate given in Figure 2 after 12 minutes is believed to 'c'in error.

b

+ 4

• af$5*A.. cY * *

,.e, y

,o,,_'

e a n.

_ mmm m, me,.m

- m m,

._ m, m M4,

s

~

The SC B 11guld level response is g nerally overpredictcd by the CRAFT calcula-tion. This again is believed to be caused by an overestimation of the auxiliary feedwater flowrate to SC B, especially between 3 and 9 minutes.

Figure 2 shows the auxiliary feedwater flow to be very; low over this time period and very stable.

This may be due to an initial instrumentation offset and no feedwater may have been delivered to the steam generator in this period.

Once a sustained auxiliary feedwater flow is established to the SG, the CRAFT calculated level increases are in reasonable agreement with the data.

Figure 4 shows the SG second'ary side pressure response during the transient.

~

f CRAFT 2 predicts the pressure response for the A loop SG reasonably.

Between 4

^

and 6 minutes, the calculated SG pressure increases above the data. Over this i

tine period, it is believed that the measured auxiliary feedwater flows are low.

This conclusion is consistent with the level comparison shown in Figure 3.

For the remainder of the transient, the prediction is higher than the measured SG prassure.

The secondary side pressure for SG B was generally underestimated throughout the transient. This is caused by condensation of the steam within the SG due to the excess auxiliary feedwater flow utilized in the calculation.

B.

Primary Syntcm Response

\\

, Figure 5 shois the A loop temperature response during the tent. The hot leg tec-perature compares well with the transient data until 13 minutes.

After this time, the CRAFT 2 calculation continues to show a decrease in the hot leg temperature due to the continued feeding of the A loop SG.

Tbc data shows a flattening of the hot leg temperature due to the control of the SG level.

This supports the belief that the auxiliary feedwater flows after 12 minutes is lower than the values indicated by Figure 2.

The calculated A loop cold leg temperature response is consis' tent with the data trend, but generally overpredicts the data after 4 minutes.

This is caused by the overprediction of the SG A secondary pressure discussed previously.

The B loop temperature response is shown in Figure 6.

Due to the overprediction in the B loop SG level and.underprediction in the SG pressure, the hot leg tem-peratures are underpredicted.

9 6

3

- g Figures 7 and 8 show the prcssurizar level and system pressure comparison.

Itand calculation 3 which were parformed ' indicate that these parameters are n.ot consis-

-tent.

Examining these. figures, it is seen that the calculated pressurizer icvel response is in good agreement out to approximately 12 minutes. Af ter 12 minutes, i

1 the continued overcooling of the A loop, due to the overestimation of feedwater flow, results in an underestimation of the pressurizer icvel.

2 The pressure response shown in Figure 8 shows that the CRAFT 2 calculation under-predicts the data..llowever, as mentioned previously, this is not unexpected as the system pressure and pressurizer, level are not consistent.

V. ' CONCLUSION' A sequential auxiliary feedwater flow transient.has been benclunarked in this analysis using the CRAFT 2 code with the 3 node SG model used in small break 1

evaluations. The site data trends were reasonably reproduced by the code.

In many cases the validity of test boundary conditions were questionable and esti-r:ates of the test data were used. However, the results provide assurance that the CRAFT 2 code is capable of reasonably predicting the primary system behavior indicated by the test if the boundary conditions were we.. defined. Thus, this study has demonstrated that, in spite of the simplicity of the CRAFT 2 steam generator model, the CRAFT 2 code can estimate, with reasonable accuracy, a tran-sicn't highly dependent on the' steam generator. Thus, the ability of the small

~

l break model to calculate the effect of steam generator heat' removal during a small break transient is reasonably assured.

O a

w e

e e

G a

e t

4 e

4

' 16. '

4 4

e e

W 6

g l

'~'

~ -

4 REFERENCES.

)'

,g

.1 -.R. A. Nedrick, J.J. Cu'dlin, and R.C. Foltz, " CRAFT 2 Fortran Program for

' Digital. Simulation of_ a Multinode Reactor Flant DurinS Loss-of-Coolant,"

BAW-10092, Rev. 2, Babcock & Wilcox, April 1975.

2 - Letter J.ll. ' Taylor - (B&U) to S. A. Varga (NRC), July 18,-1978.

4 9

A e

2 e

e e

, D e

j e-1 l

O I

e s

+

4 sm 9

e J

t 9

5 Y

a e y e

5 6

t-O e

• *

• 171. +

=

e a

u

+

s y

y...

-y..

.,, -. - -,,,.~,.,,,,,, -, - - - +

-y-.

• t a

,1..I I C..,,,,Ja.i... G..,0..... L,,,..,

G G

,. Od h @,,

6 T

e1 n

A it 4

I

=

0 P LL

.6

(

(

8 o

[e.

,e 0

e,g i.e

• I *'

e'

• d >> O T

e e

r i

G

"" - 6 e

O h3f(5 A00lflChAL ttfA haf sesa C4 OlaGtiv lats!

l

@ 45 CC4f alhPtti MCI II heat 3

PA fit

.t1 LE AR FifM FRC4 f t!!E f3 CC4fil%'t%i LPl 2

@ 15 plfttu Ltan Pain FEC CaTA4hn'.T IC E2(as noct PAIN PA ffl t!PRt$ttf1 CGttalhetti $PR47 Stifte h001 h3 10(EtiPitiflCM PAfil p$.

IDInf tflCallC4 6

0C:4C08t4 l.2 CCat 2

LCatt PLErJs 3,4,18 it hof, LIG PIPihG 3

tCat 5,20,40,42 1101 LEG. UPPER 4 le

1. 0! 115 PIPl%G 4,21 3G ft!!5 S tl 15 L DPPit W183 T,22

$G LCuta king 6,16 Effle Gthf R4f 3A II,8t!

I CCRE 87P81$

I.,17 IICC408tf 15 9,13,24 COLO LLC PIPl4G 8,18

$$ LCt!E NERG 88.84,25

, PURPS

~

1,11,19 COL 8 LIG Pitlht 11,02,65.16.24,27 COLO LIG PIPihG 18,12,20 C0lO LIG PIrint 17.31 GCantecia

- F 13,

DPPER B084CCatt 23 LPt il pat $3utilla 23.23 GPPit $Cas,gete e

27 C04T Al%8t h!

33 PR(35UAlltR Il DPrtt plt %UE 32

%!kt V&Llt 24,25

$$ WPPit DI43 33,34 Llat & Pitilag Pats 33,36 IIPI,

3F C04falnatti 5 Posts 41,43 3G CPPER klag 4

'(

=

G e

L

i FIGURE 2.

STARTUP FEEDWATER FLO.W

- l l

1

~ A LOOP TEST DATA 1

l

__ B LOOP TEST DATA t

4 J

5..

\\

,s

/

/

Q

/

/

/

r-4..

I l

LD o

I x

I l

s.

3 I

.c sr-I m

a l

l x

I o

E l

2

.1 I

l

+-

1 r

1 r

=

l1

/

\\

/

i.

I1 A

I \\ l\\

o I

l 1/ \\

/

1 1

11

\\

Y L _ _ _ _ _ _ _ s'/

\\

/)

~

l l'

I O

0 4

8 '

12 16 20 i

Time, Min 43 6 A'h'.

, ~, ' 'Y " a4 I' S

~

1. Yidi*'s

[i

' * ~

4

1. , b Y-[ega '

mb-e J

MN I 4-h' ENW R EM mop [f.4 P>N/ -Q w

i!

FIGURE 3 STEA!,1 GENERATOR LIQUID LEVEL (TEMPERATURE ADJUSTED)

l 2

5 i

1 A LOOP TEST DATA A LOOP CALC. DATA

---

B LOOP TEST DATA B LOOP CALC. DATA 25

./

f 20

[

~

l

/

.l' 8

u

/

O 15 5

./

m

~

/

w 5

/

/

l.

2

/

/

x a

/

8

/

10 --

l

_/

,/

~

Y

/

./

o

/

/

./

i/

m EE

/

[

5

/

/

/

'/

/.-!

m n.

/

t i'

/

\\

a

^;

l

/

g

/.

i

,1

/

~ ~,

i.

ss 0

D 4_

.8'"

12 16 20 l

Time af ter loss of power. Min i

I FIGURE 4 S.G. SECONDARY SIDE PRESSURE i

i A LOOP TEST DATA A LOOP CALC. DATA

^^

~---~

1100 -

=- B LOOP JALC. ?ATA l

1000 -p' s,N I

I.

s N

i g

\\

d

\\

5 S00 --

b\\

s N

l g-N y.

xs se 800 -

's N \\

x/\\

m E

\\

.\\

N\\

\\._/

N.

!\\

g

\\

\\

\\

700 --

\\.

\\

g

'\\

i

~%. ?'

'('%.% s,

\\

\\

500--

\\

\\

l

\\

\\

i

\\

\\

I; i l 500

'I t

0 4

'8 12 10 20 i!

Time, min

_21_

' FIGURE 5 PRIMARY A LOOP TEldPERATURE t

i LOOP A HOT TEST DATA I

' LOOP A HDT CALC. DATA

- ----- LOOP A COLO TEST DATA

--- - - LOOP A CO'LO CALC. DATA i

i e

S

\\

m 580 -

N \\

I N

i

'g N

\\

\\

570 --

i

\\

N gi x..

c\\..

x s

1 e

500 --

x 1

B

\\\\

'N\\

?

\\

k.

\\

N I

t

\\

550 --

.)

\\

\\

\\g\\

\\

g

\\

N.

.\\

\\

\\

i

\\-

540 --

\\

\\

\\

\\

s A

N.

N

\\

\\.

\\

i.

'\\

'530 --

\\

\\,

\\.

s

\\

s

\\

N g

x s

520 0

4 8

12 16 20 l

Time, min I -

FIGURE G.

PRIMARY B LOOP TE!!PERATURE LOOP B HOT TEST DATA

/

LOOP B 110T CALC. DATA

/

~

LOOP O COLD TEST DATA

.- LOOP B COLD CALC., ATA K

580 --

~x s\\.\\

\\

-I 570 --\\'

%_\\ -

x s

h

\\

\\

\\

'N N

\\% N.N

)N

\\.

w3x m

560 --

g\\

\\

'(

\\, \\g

\\

\\

g Ite

\\.

H 550 --

\\'\\

\\\\/

-- i 540

\\\\

530..

\\

\\

.)

520 0

4 8

12 16 20 Time, Hin e.

1-4-30 e

FIGURE 7.

REACTOR VESSEL PRESSURE I

TEST DATA e

=X XX X 'A X CALCULATED DATA 2200..

I.

$I 2100 -

i.

e N

g, 2000.-

9 O

L U

M e

D 6

1900..

+

o CC n

• h y *\\

r I

f A

1800..

\\+

l 3

t t

I e

s

^

e

. 16 20 I

0 4

8 12 1

i.

Time, Hin 4

f 14

.?

2 I

o.

l,

)

t

+

8

?[TN7T**-'F'} W.g *M*#9 * *M 2 L' >*% f, Wrt-" =*", p f W ger eMy Wg

+,

-, w

, Mim. J%,..

n

~,.i_m.+-+-

r

(

Ma l k.. -. i.m-.

c,s%g d r. -

^?

-I T. *'

j. 7;'.

2 4.,;

}

[*

g.2 y.

.,=

d 4

g.

~

(

FIGURE

0. REACTOR VESSEL PRESSURE TEST DATA

.e c

a CALCULATED DATA

.e d

2200..

a- -

.\\

\\

~

\\

\\

2l00

--I

'~

\\

N N\\

a.

s 2000..

N a

g N N E,

N s N

o so N

o

\\

u c.

N 1900..

s o

s M

\\\\

N

\\\\-N

.e 4

1000 s

N

\\

O tl 8

12 16 20 Tirac, 1lin O

e W

L' 1_ A _ o r)

9 4

1 Item No. 7: -(Attachment A to Encleure 1 of Reference 1) lB. ' Provide ~ justification of relief-and safety-valve flow models used in the CRAFT 2 code.

i~

RESPONSE

1

. The CRAFT 2 code, which is documented in topical report BAW-10092, Rev. 2,

does not have any special models for prediction of the fluid discharge through the relief and safety valves..Rather, they are modeled as leak paths from-the pressurizer control volume to the containment. Thus, the Bernoulli (orifice) equation is used for subcooled discharge, while the Moody correlation is used for saturated steam or two-phase discharge.

Th'ese models'are the same as those used in B&W's ECCS Evaluation Model.2

- Since little information exists'on the flowrate through pressurizer valves

. for subcooled or two-phase fluid conditions, it is impossible to ascertain

- the accuracy of this medeling technique.

Since pressurizer leaks are in-herently less severe than the breaks in~the cold leg pump discharge piping analyzed to demonstrate compliance to 10 CFR 50.46, a truly realistic model:for'the discharge rates is not necessary. However, the modeling L

technique' utilized is expected to reasonably approximate the discharge rat'es and their' subsequent effect on the RCS.

J System response to relief valve actuation have been analyzed and submitted I

' to the Staff in Section 6 of the May 7,1979, report. - The cases speci-3 fically analyzed'were:

E 1.. A loss of main feedwater' accident which rcsults in actuation and a

~

subsequent sticking open.of the pressurizer relief valve was addressed.

Offsite power was assumed to remain available and only one HPI train was used for emergency core cooling. This analysis is similar to the

~

JTMI-2 event that occurred'on-March 28, 1979, and demonstrated that,-

if one HPI pump remained available, no core uncovery would have oc-1

- curred. --

2.

ll stuck'open.PORV assuming a-loss of offsite power and only one HPI train available was-analyzed.- Results of this evaluation demonstrated n

that core uncovery would also not occur.

2.

._m w a, e.

s r-

-+,

v--,

,m

f

%]

r

An additional analysis of. the effect of;a pressuriser break which supple-mented those' presented in reference-3, snts provided to the Staff in a 4

letter from J.H.-Taylor.(B&W) to R.J. Mattson (NRC) dated May 12, 1979.

' That analysis examined the effect of the stuck-open PORV case, Case 2 above, except-the auxiliary.fecdwater system was assumed inoperable. The.results of'that evaluation showed that, even without auxiliary feedwater, one HPI

-Pump can handle the' accident provided that realistic decay heat values are utilized. In all of ~ these evaluations, the 10RV was modeled via a leak path' representation in the CRAF12 code.

The orifice area of the PORV was modeled as the leak area (1.05 in.2) and a discharge coefficient of 1.0 was utilized.

Tho' method for modeling the PORV described above does result in a pre-dicted steam flowrate, at the valve rated pressure, which is in excess of the design (rated) flowrate. An alternative modeling approach is to use a discharge coefficient (C ) which, at the valve rated pressure, would D

yield the valve rated flowrate. For the 177-FA plants, this is a C D

approximately 0.85.

For the first two cases described above, this model-ling approach would result in a slower system depressurization and a slower discharge of the RCS inventory. Thus, the use of a C

.0 used

=

D in' previous evaluations results is a conservative assessment of the tran-sient.- For the third case, the use of a smaller C w uld result in a D

larger repressurization following the loss of the SG as a heat sink and i

the change in the discharge from steam to two-phase flow. However, use of a C f 0.85 would result in an inventory loss less than that calcu-D lated in reference 4 and no core uncovery would occur.

Besides the cases involving actuation of the. pressurizer relief valves, analyses were performed for a total loss of SG heat sink and are provided in references 5 and 6.

In those evaluations, the pressurizer safety valves were exercised.

To model.the~se valves, the leak path representa-tion-was used with the leak path opening and closing at the opening set-

~ point of the _ valve. _The valve area and C was chosen such that the rated D

' flowrate for the valve would be simulated at the valve rated pressure.

Because of the large relief capacity of the valve, the system pressure oscillated;within a few psi of.the valve setpoint and the valve was.

4 cxercised intermittently. 'Thus,1any discrepancies between the modeled

=-27,

g.~

e

---e

2nd the' actual relief capacity of the Jressurizer safety valve is not ex-

. pected to significantly alter the system response.

While'there is little information availabic on the discharge rates through-the' pressurizer: valves, it is also important to note the breaks in the pressurizer are bounded by breaks in the cold leg pump discharge piping.

Pump discharge breaks are analyzed to show conformance of the ECCS to meet the criteria of 10 CFR 50.46.

The reason that cold leg breaks bound breaks-in the pressurizer was discussed in detail in reference 3.

Therefore, it is not necessary to simulate the actual relief capacities of the pressuri-zcr valves in order to' demonstrate the ability of the ECCS to mitigate the consequences of a loss of RCS inventory through the valves within the criteria of 10 CFR 50.46.

REFERENCES 1

BAW-10092, Rev. 2, " CRAFT 2 - FORTRAN Program for Digital Simulation of a Multinode Reactor Plant During LOCA," R.A. Hedrick, J.J. Cudlin, and R.C. Foltz, April 1975.

2 BAW-10104, Rev. 3, "B&W's ECCS Evaluation Model," B.M. Dun, et al.,

August 1977.

3 Letter J.H. Taylor (B&W) to R.J. Mattson, Fby 7,1979, " Evaluation of Transient Behavior and Small Reactor Coolant System Breaks in the 177-FA Plant."

3 Letter J.H. Taylor (B&W) to R.J. Mattson, May 12, 1979.

P Letter from R.B. Davis to 177 Owner's Group,

Subject:

" Complete Loss

',of Feedwater Transient," September 11, 1979.

)

- 16 ' Letter from R.B. Davis to Mr. C.R. Domeck,

Subject:

" Complete Loss of Feedwater Transient on Davis-Besse," September 11, 1979.

e,,

pA A --. s -..

J Item Fo. 7:

(Attachment 1 to Enclosure 3 of Reference.1) 2A.

Provide justification that. the 3 node steam generator model used in the CRAFT 2 analysis of small breaks is adequate for the prediction of steam generator heat transfer.

RESPONSE

The B&W ECCS Evaluation todell for small breaks utilizes a three-node repre-sentation, in the CRAFT 2 simulation, for the prediction of steam generator heat transfer following a small break. Two of the nodes, stacked verti-cally, are used to model the primary side of the once through steam genera-tor (OTSG). The upper node includes the hot leg piping, from the center on the 180* U-bend at the top of the vertical section of the hot leg to the SG upper head, the upper head of the SG, and the upper one-half of the tube region. The lower node simulates the lower one-half of the tube region. The third node is used to model the secondary side of the OTSG.

To evaluate the suitability of this modeling technique, the unique charac-teristics of the OTSG and its effects on the small break transient must be examined. As is shown later, for small breaks evaluated with the auxiliary feedwater system operabic, heat removal via the SG is not necessary for the worst case breaks, 1.c..

those that result in core uncovery, in order to successfully mitigate the transient.

For the smaller. breaks, heat re-moval via the SG is necessary. The three-node representation utilized appropriately models the heat transfer characteristics of the OTSG. For the smaller breaks, heat removal via the steam generator is necessary and the heat transfer characteristics of the OTSG must be appropriately con-sidered. Although the 3-node SG model does not rigorously account for I

the heat transfer process-that will occur, it does provide a reasonable-4 representation of the effects of these heat transfer processes in the OTSG.

Since these smaller breaks exhibit large margin to core uncovery, the CRAFT 2 SG model is adequate for demonstrating compliance to 10 CFR 50.46.

2 l[n performing small break cvaluations, the CRAFT 2 code-is used to pre-

dict the hydrodynamic response of the primary system including the effect e

e,

n

of SG heat transfer during the ' transient. The option 2 SG model, which is explained in detail in Section 2.6 of topical report BAW-10092, Rev. 2, is utilized to predict heat flow in the SG.

The calculation progresses basically as follows:

-1.

Based upon the initial steady-state heat transfer characteristics of the OTSG and the initial primary and secondary fluid temperatures, an overall UA for cach region of the SG is calculated.

2.

The calculated steady-state UA can be modified by user-specified in-put options. These include an input multiplier table versus time, multiplied based on the primary side control volume mixture height during the transient, and a multiplicr for reverse heat transfer, i.e.,

heat flow'from the secondary to the primary side of the SG.

3.

Using the modified UA and the calculated primary and secondary side control volume temperatures, the amount of heat transf2rred is calcu-lated.

In performing the small leak calculations for demonstrating compliance to 10 CFR 50.46 for the operating B&W plants, no input multiplier versus time is utilized, nor is the modification based on primary side mixture level used. However, a multiplier for reverse heat transfer of 0.1 is u tilized'. This multiplier and its basis is explained in the ECCS evalu-l ation model topical report and is uti.lized to reflect the change in heat transfer regime on the secondary side of the SG for reverse heat flow.

The.0TSG design of the B&W designed operating NSSs allows use of a simplis-tic model for calculation of SG performance during a small LOCA transient.

With the loss-of-offsite power, assumed in design calculations for small breaks, and the subsequent loss of main feedwater, the auxiliary feed-watersystemisactuatedandwillbecomeoperableinapproximately(ff) seconds.and control the secondary side level. The auxiliary feedwater enters the SG very high, approximately 2 feet below the upper SG tube chect, and is. sprayed onto the tube bundles. Thus, heat transfer will occur in the upper portion of the SG independent of the actual icvel in the SG.

The introduction of auxiliary feedwater to the SG has two ef-fccts on the small LOCA transient.. First, it raises the thermal center in the SG during the natu'ral circulation phase of the accident which

-30,

results in afcontinuation'of circulation through the RCS, for some period

= of-time, even while inventory is lost-from the primary syctem. Later in the transient, after sufficient inventory has been lost from the system, circulation will be interrupted and the auxiliary feedwater, for a certain range of.small breaks,.will condense steam on the primary side of the SG; thereby maintaining the primary system pressure near the secondary side

~

The analytical approach utilized for the small break evaluation pressure.

- is consistent'with this performance of the auxiliary feedwater system.

- It should be noted that between the time that circulation through the loops is lost and the time that the primary side SG Icycl has dropped to-the point where condensation heat transfer will occur, system repressuri-zation can occur as heat removal via the SG will be lost. This phenomena occurs only for the very small sized small breaks in which the SG heat removal is necessary.

If simulation of this repressurization phenomena of the very small breaks is desired, an additional node would be needed in the small break model in order to separate the hot leg and SG upper plenum volumes from the tube region..This will allow steam ;o accumulate in the upper regions of the RCS without being affected by heat removal that occurs in the steam generator.

In the analyses presented in reference 4

t 5 for-these smaller sized breaks, a model which included tho additional node was~ utilized and showed that the repressurization phenomena does

- not result in core uncoscry.

It is also important to note the role of the SG on the small break tran-sient in order to evaluate the appropriateness of the SG model utilized in sum 11 break evaluations. Licensing calculations for the operating B&W units have previously.been submitted to the Staff in references 3 and 4.

These evaluations have shown that the worst case small breaks, i.e., breaks 2

I.

which renuit in core uncovery, occur for breaks in excess of 0.05 ft. As s

i demonstrated.in the May 7, 1979 report, SG heat removal is not necessary

~for breaks of this size. _For smaller breaks, SG heat removal is necessary as the breakfalone is not sufficient to remove enough fluid volume and cncrgy to depressurize the RCS. However, as demonstrated in reference 5, these breaks are of no consequence as the SG heat removal and the slower discharge rate for these breaks easily prevents core uncovery.

i a

4-f r-3 q

,c, y.

,. g r..

,n_,

m-.-3--

^

1As demonstrated, the SG model utilized in the small_ break evaluations for the operating plants appropriately accounts'for the effect of the spatial t heat': removal processes that vill occur in 'the OTSG during.a small break.

It'was also'shcwn that the SG performance is not important for the' worse case small breaks. Thus, the CRAIT2 SG model is adequate for demonstrating compliance of.the ECCS to 10 CFR 50.46.

REFERENCES

-1 "B&W's ECCS' Evaluation Model," BAW-10104, Rev. 3, Babcock & Wilcox, August 1977.

2 R.A. Hedrick, J.J. Cudlin, and R.C. Foltz, " CRAFT 2 - Fortran Program l

for Digital Simulation of a Multinode Reactor Plant During Loss-of-Coolant," BAW-10092, Rev. 2, Babcock & Wilcox, April 1975.

3 Letter,.J.ll. Taylor (B&W) to S. A. Varga (NRC), July 18, 1979.

j 4 "Multinode Analysis of Small Breaks for B&W's 177-Fuel Assembly Nuclear Plants With Raised Loop Arrangement and Internals Vent Valves,"

BAW-10075A, Rev. 1, Babcock & Wilcox, March 1976.

5 " Evaluation of Transient Behavior and Small Reactor Coolant System Breaks in the 177-Fuel Assembly Plant," Babcock & Wilcox, transuitted via letter from J.H. Taylor-to R.J. Mattson,. dated May 7, 1979.

J t

4 4

, ^~'

(

7 1

n o

.t n

n ho w

d e

t t 6

3 a e.

a 1

o II 1

h on P t ri n

S Fa k e t

t a n s

hn en

1 t o

'l I.

^

aC e e I'

P nt d a

o r no r

a t

kt u o ';

a t C D :e ek e

k 1 r C

La R ma s a e

oc n

S sr sr r c

2.

I iB

! FL

,i a t r

! g X

a 3

ff i

O J. D f

(!

1 6

l',

I 2

i I_

2 l

e h

N C

t II P

I#

t d

t 0

2 io a

~ r r>'

e, tne 1

i s

na r

a 3

2 2

T 2

I M

'3 T

1 r

T.

o R

f

)1 e

U 3

/

G a

T

)-

r F

w z_

I a

C F

r 5

g

<9 i

a i

b 3

D 1

.< I n

g i

i.

s d

o N

?)

T i(

r 2

F

,l 5,

9 I

A R

1.e3 6

C 0

2 lV 5

l

~

I O

PL

_g-1 4

i 2

l I

8 l

I

2000

. FIGURE 2 SYSTEM PRESSURE VS TIME h

2400 fi T!al-2 DATA v

i CRAFT-2 SIMUL ATl011

}

l 2200

- 't I

t 1

I 2000 l

\\

2

\\

E

\\

at

\\

1800

.g 2

\\

E

\\

N N

N 1000 1400 1200 4

1000 I

I I

I O

1 2

3 4

5 I

Time, IJinutes.

s

• se 0

j 4

4 1

0 0

i 0

me, 8

N O

I 0

T T

2 A

N 0

1 L

E 0

U D

M I

i 0

e 7

I C

S C

A n

e 2

N T

2 F

s A

I 0

R M

0 i

C T

~

i 0

0 0

6 1

E

. ^ -

RU S

S s

0 e

E 0 s t

R i

0 d u

5 n P

n 0

a o

i c

T M 8

'e N

S A

e L

m O

i em O

T 0 i C

0 T R

e' 0

4 C.

6 i

w O

T.

0 A

ER 0

0 0

3 3

E RU I

0 i

G F

4 0

0

, 0 2

w 0

i 0

2 y

\\

, 0 0

g 1

s N

\\

4/

\\

r O

f7 r 0

0 0

0 0

0 0

0 0

0 0

0 0

5 0

5 0

5 3

2 2

1 1

n~ c. d

~

. bT' i {7r4vi::%

.'hn l

I, gU s. bis'/ tW'AmM?

i H '

Y 'i:Em..

4;i?

i.' I i v.. gI ( ,E k; iyI

w 4 e s,*3 9 I / . n ( 1 Mu 1 ~ e a' xm n2' I ~ G O W ""5 - l / - n= I l', ci ;- E, >K e,r1 f O EsJ 9-- Er1 { l CD 4 M ksJ "s W M yn~ l N S$"- S -m ~ 2 M o m - - i - i j . j 1 g uj i l I = I a: M o l I I uJ i s-7 l G ~' h 1 -h s my .g' f a ~ w 1 g i l bl O' s D I ( c> 88 3 o d + i' H b. l-1 i l j = l I a. [ - = i l I I \\- I - n 1 g I d i ') / I -) I - = t\\ s % ~ * *'* m & - -.~ ~. -~~-'-w..s e a. e i e E. E,,, E,. E., E E n ul 'j9A31 -36 9 e -.-q

4 FIGURE 5. HOT LEG FL0rl t i e LOOP A HDT LEG 20000 a' --- LOOP B HOT LEG \\ - - LOOP A-T!41 OATA ~

---

-- LOOP B-Tid! OATA -

\\.._,. s..... I^ 16000 1 j

w. s, g-N.e:3 t

+...y'...., ( i N '., ~ \\,f:." s i. 12000 t s g-q \\- q s, -E F N) t 1 ' y ~ 1: \\. i 8000 o u 1: \\ ~ 1. il. 1 i ti. \\. 4000 i'.. L \\. t: h \\-\\. I: 1 II \\ C y 0 0 20 40-60 80 100 120 140 Time, Minutes 8 0

0 0 4 0 \\ 1 0 8 s -[././ s s' jl i .!:iI* *

• l.'
• l. iI.! i :i1 :l.i;.::.

5 l/ .j*l l

j:/

l s f.. ) l . (- f l:l I Ilt j/ / / I,1 0 i .,., a. .j/ 0 / / /ji r 2 I j.\\ g 1 0 ,l I' i 0 .Nf - 7 / ? v.. ~.s\\.9* %:. 0 0 \\%.41-N.. 0 0 i 0 1 1\\ Qt 6 \\ .h ))~.5 s,- ell'% A A E T T g.w i 0 R 0 A A U 0 D D i T 0 A 5 2 2 8 R s I i 2 2 s f= 4 E d M t T T t e F n T T F F t M o A A u c R R R R ~ n E e N N C C S s. i T M y" s A S B A G 0 P P P P E 0 i e L L L L e 0 m m ys L T T I T 4 i i 0 T T O 0 0 D T H H H H. f]h, i 8 0 T T T T. h 6 .~ 0 d 0 i a '0 3 u w 4 b l 0 F . :=. 0 '0 i 0 2 i.1.Ys% 0 i 2 s '0 I

• . f. *.

1. i. :.g\\i 0

0 i y-1 \\;li \\ 1.'- '..: !. l M m !E~ ~ k?'- 0 0 0 0 0 0 0 0 2 0 8 6 4 2 0 6 6 5 5 0 5 5 iy

~ FIGURE 7 COLD LEG TENPERATURE Time, Seconds 1000 2000 3000 4000 5000, 6000 7000 660 i i i i i TCOLD LOOP A WR TMl-2 0 ATA

---

TCOLO. LOOP B ViR TMI-2 DATA 620

___.._....TC0'.0 LOOP B CRAFT 2 '~ TCOLO LOOP A CRAFT 2 '\\. 58'O ' ! s\\' \\, ~ .s .o \\ 6 's \\. (%T 'C#... 5',_,~.,m m. .w -> " '" C -p m \\,s\\ y - ~ ~ ~ ~ _ ~ ~ ~ ~ ~. a.g 540 Q ~ - - % ;,eq %,p"w rr A n N\\ 500 \\ N \\ s 460 i I i i i r 0 20 40 60 80 100 120 Time, Minutes l

9 12.0 _ FIGURE 8 CRAFT 2-CORE MIXTURE LEVEL 10.0 + 8.0 O l 8 3 8 6.0 l 3 x I i e 4.0 e e 2.0 0.0 I I I ^ 4000 5000 6000 7000 8000 O.0 1000 2000 3000 Time, Seconds

r-9 l - c-. n . ~ a o 84 e, -:o J G. a Of n. o W y ~r hJ n m = s m ~ (4) LO = -= O,_ Wn s s '.P .e-3 e e Y n .4 H a. n h ~ m s o .N a Dj a* z w tae o I x W .x. u = g >-- +- oe m . a4, _ c

• = -

m 5af e,= tsJ o e3 o H = u. a a o e e f I t i a a o c; i: E a n 0 n 5 'sD!oA E435AS 33 ' i J

Item I'o. 7: (Attachment A to Enclosure 1 of Reference 1) I 2B. Provide the reactor system response to a struck open PORV for the case of a small break which causes the reactor sys-tem to ptessurize to the PORV setpoint.

Response

See Attached.

~J-3..,_.__..........- SMAl,L BREAl* WITil FAILED PORV ~ j 1. INTRODUCTION It has been established in reference 1, thht very small cold leg breaks (<0.01) will reprcosurize to the PORV setpoint of 2465 psia if the auxiliary feedwater Since there is a. probability of the PORV sticking is deltyed significantly. open af ter being' actuated, concerns have been raised regarding the impact of This report presents the results of an analysis this consequential failure. cold leg break-with the subsequent failure of the PORV to close. 2 of'a 0.01 ft 2. SU? DIARY & CONCLUSTONS t' .~ As has been ocmonstrated by the analyses presented in Section 6 of reference 1, small breaks _in the primary system will not cause a repressurization to the Under this PORV setpoint unless all feedwater is lost to the steam generators. 2 situation, there exists a class of very small breaks, (less than 0.01 ft ) An analysis is pre-wherein the system will repressurize to the PORV setpoint. cented herein for a 0.01 ft break, without feedwater to the steam generator, 2 At 20 which results in a repressurization to approximately the PORV setpoint. minutes, the PORV was actuated and was assumed to stick open. As is demonstrated in Section-4,'for the 177-FA lowered-loop plants, operator action by 20 minutes to manually actuate the two high pressure injection trains will kcep the core covered. A qualitative analysis is also presentec1 which demonstrates that reestablishment of auxiliary feedwater by 20 minutc s, for both the 177-FA raised and lowered loop plants, will prevent core uncovery. break with no auxiliary feedwater can be mitigated safely 2 Therefore, a 0.01.ft with D&W's present operator guidelines. These operator guidelines require ~ if the AFU ' establishing feedwater"to the steam generator as soon as possible, is not available initially, and manual initiation of the IIPI' upon loss of the ' steam' generator heat sink or caturated conditions in the primary system. 3. IICTHOD OF ANAll/ SIS Evaluations of very small~ breaks which result in repressurization phenomena are presented in reference 1..These analyses demonstrate that if auxiliary fecdwa'tcr is delivered to the steam generators, the primary system would not ~ ropressurire to the PORV setpoint. However, the analyses in reference 1 also e.

v demonstrate that if feedwater is not delivered to the steam generator within 20 minutes, there is a class of very small. breaks, 1 css than 0.01 ft2, which will Since the PORV might - result in system repressurization to the PORV setpoint. stick open af ter being actuated, concerns have been raised regarding the impact ~ of this consequential failure, j ' An analysis of a 0.01 ft break in the cold icg pump ditcharge piping, without 2 i auxiliary fecdwater to the SC, was performed wherein the PORV was actuated and assumed to stick open. As has been demonstrated in reference 1, larger breaks will result du automatic actuation of the HPI system and will not repressurize. i' While smaller breaks will repressuri c to the PORV setpoint earlier, less in-1 2 rmall break with ventory would be lost out the break. Therefore, the 0.01 ft the subsequent failure of the PORV is expected to be the worst case for tran-Eicnts of this type. The analysis was performed using the B&W ECCS cvaluation model for the 177-FA -lowered-loop plants.2 The analysis was performed using the same model and 4 assumptions listed in Section 6.2.1.3.5 of reference 1 with the only changes being those made to reficct the PORV sticking open. Key assumptions of the r I-1 analysis are listed below. 1. The initial core power Icyc1 is 102% of 2772 Mh't. 2. The core decay heat is based on 1.2. times the ANS standard. 3. Operator action was taken at 20 minutes to manually actuate both HPI pumps. f 4. The PORV was modeled as a leak path on th-top of the pressurizer. The. 2 was used, however, a C f 0.72 was utili cd in ' orifice area of.0073 ft D order to reficct the proper relief characteristics of the PORV with the } 4 Moody critical flow model. ~ ,5. -The PORV uas opened at 20 minutes. This is consistent with the operator Euidelines for a LO'CA with no feedwater to the steam generators.

However, if the operator had not acted within this time frame, approximately a 2 ninute delay in operator action would have resulted in the PORV being l

) J actuated automatically. ,,e 4, _e ,.-m., w 9p.. 3 9. , m9 e a

4. RESUf,TS Figurcs 1 through 7 show the system response during the transient and Table 1 presents a sequence of events for this accident. The resultant system pressure response of a 0.01 ft2 cold icg break with no AFW is shown in Figure 1. This particular response is due to (1) the Joss of the SG heat sink; (2) no automatic IIPI actuation prior to the loss of the stcan generator heat sink; and (3) the opening of the PORV and actuation of the HPI at 20 minutes. As seen in Figure 1, the pres'surc' initially decreases following the break opening. During this de- 'h' j pressurination period, the reactor trips, the pumps trip, the pressurizer captics, and the stcan generator secondary inventory boils off. With the loss of the SG 'f ~ heat a k, the primary systen starts to repressurize before the ESFAS signal is,.v'd n . ()/ y .i reached. Therefore, the !!PI is not automatically actuated. The system repres I surines to 2350 psia by 20 minutes at which time the PORV was assu=cd to open. ' .7 This is only 115 psi below the PORV setpoint which would have been reached ap-proximately 2 minutes later. Ilowever, the operator is instructed to manually i open the PORV if the system repressurines and the SG heat sink is lost.

Thus, the opening at 20 minutes is not totally arbitrary.

During th2 system repres-surination the pressuriner 1cvel increases (Figure 2) and when : he PORV is opened e the pressurizer rapidly fills wi'.h two phase nixture. At the tim, nf the PORV opening, the two llPI pumps are manually' actuated, and due to the addition of the cold nakeup water and the additional leak path area, the RCS depressurines. The inner vessel mixture height is shown on Figure 3. As can be seen, operator action by 20 minutes to manually actuate the IIPI prevents core uncovery and n minimum two-phase mixture Icvel of fc t above the top of the core is main- ./ tained. Long term cooling is established at 25 minutes as the injected HPI fluid exceeds the core boil-off. Thus, the acceptance criteria of 10 CFR 50.46 are satisfied. While the analysis performed herein addressed the effect of operator action to k manually actuate the HPI by 20 minutes, the effect of operator action to manually restore the auxiliary feedwater within 20 minuucs can be qualitatively assessed. I As has been shown in Section 6.2.1.3.5 of reference 1, actuation of the auxiliary feedwater system at 20 minutes for a 0.01 ft2 break results in a rapid system depressurization and the subsequent actuation of the HPI. For the case analyzed 'herein, the depressurization effect of the auxiliary feedwa;.ar would be faster than that shown in reference 1 due to the effect of the loss of inventory through the PORV. Thus, the llPI would be actuated earlier and long tern cooling.

vould be establir.hed faster than that shown in reference 1. Therefore, no core uncovery is expected if the operator only actuates the auxiliary feedvater sys- s tett within 20 minutes and, contrary to the guidelines, does not manually actuate the llPI. O e 4 e e 4 i A e O O O S e .1

r I: I / Tab)c 1. Scquence of Events Event Tinc, a 0.0 1. 0.01 ft2 cold Icg bre.' occurs 54.5 2. Reactor trip, loss of Iceduater, and RC pump trip 3. !!ain feedwater coastdown ends 60.0 270.0

4.,S0 secondary boils dry 1200.0

'5. PORV opened 6.< 11PI is manually initiated 1200.0 7. Long term cooling established 1510.0 4 4 e g =m--- e e l 1 ) 6 -47

f. RETT.RE!!CES Letter J.!!. Taylor (BM1) to S. A. Varga (f;RC), " Evaluation of Transient I Behavior and Snall Reactor Coolant System Breaks in the 177-Fuel Assernly I'lant," May 7, 1979. B.!!. Dunn, et. al, "BM1's ECCS Evaluation Model," BAW-10104, Rev. 3, Babcock 2 6 U11cox,'May 1975. w S .48-

f N.,, Figure 1 ? COLO LEG DREAK W/I'0 AFW 2 HPI'.S & STUCK PORV .01 f! AT 20 tilli. - N00E 14 PRESSURE VS Til,1E 2400 ~ 2200

.2 2000

= E s ^ 1800 1000 1400 1200 t O 500 1000 1500 -2000 2500 Time (sec) .49-

1. -

I i / 4. e figure 2 .01 FT COLD LEG DREAK Yl/t;0 AFYI 2 HPI'S & STUCK 2 PORY AT 20 Miti. - PRESSURIZER LIQUID LEVEL r 00.000 .._e_. 50.000 1 A 40 000 N ~ c b 3 .30.000 3 4 20.000 ee-9 10.000 = = e* menn e o me m b-I 0 0 500 1000 1500 2000 2500 Tirae (sec) J'-

• w 4-v

Fif.ute 3 01 II2 COLD LEG BREAK Ti/N0 Afil 2 HPI's to STUCK PORY AT 201.11H. - UPPER PLENU:.1 LIQUID LEVEL ~18.000 s i -] 10.000 l .r 14.000 l \\ O 12.000 Ei ~. =. e'g 10.000 3 y 8.000 \\ 0.000 Jl(1 2 f: 8 3 4.000 -L f,.-y 7 -f \\ u J N4)'\\ kN f 2.000 O 500 1000 1500 2000 2500 Tine'(sec)

f. i / o I Figure 4 l' 2 .01 FT C01.0 LEG BREAK W/NO AFVI 2 HPI'S & STUCK PORY I AT 20 1,llN. - PORY LEAK FLOYl l 60.000 I i e l l .50.000 ~ f l ~ () 40.000 8< e 3 30.000 u. N 20.000 10.000 0 t i i i 0 .500 1000 1500 2000 2500 Time (sec) ' 52-

^ .,./. ,i s .t l / l s Figure 4 i I 2 .01 FT COLD LEG BREAK WND AFVI 2 HPI'S & STUCK PORV l AT 20 f.llN. - PORY LE AK FLOYl l 30.000 ~ i t .50.000 I ,q, t i 40.000 I$ b 30.000 o u. N l 20.000 10.000 0 0 500 1000 1500 2000 2500 Time (sec) 52-

/ pe 9 fintire 5 .01 FT2 COLO LEC BRE!K WMi0 AfW 2 IIPI'S & ST'JCK PORY AT 20111N. - PORY LEAK FLOW QUAllTY 1.200 i 1.000 i 1 . n. .000 u E a n2 .000 ~ .400 .200 O! 0 500 100'0. 1500 2000 2500 Tine (sec) '

/ Figure 0 2 COLO LEG BREAK U/ tin AFW 2 HPI'S & STUCK PORY 01 FT AT 20 Mlit. - COLO LEO OREAK FLOW 35.000 e 30.000 l 25.000 i ^ i ,c ,_g j + -, 5 20.000 l S I ~C2 dc 15.000 10.000 1 q' l ,l} - f,jlJ Q i 1 1 l li i 5.000 I ll L L" e .e 0 l 0 500 '1000 1500 M N Tine (sec) s e

V _ l...,. ~ ~ Figure 7 .01 FT2 COLD LEG BREAK Vi/NO Afi7 2 HPl'S & STUCK PORY AT 20 IJIN. - COLD LEG BREAK LEAK FLOW QUAllTY I i. 1.200 j 1.000 s I .a ) .000 D g. 'i f a g,i i !!j'- c' f .I l i E .000 l l l r t I l i. l li I l l'l I i .400 ~ Hp i j! l l t 4-lkl1,ffD ! 1 I j K U .200 l f I i O ^ l f 0 500 1000 1500 2000 2500 Tic:e (sec)

.~ -y>

4. '

/ - would be established' faster than that shown in reference 1. Therefore, no core r ~~.. - -.uncovery is expected -if the operator only actuates the auxiliary feedwater sys-tem within 20 minutes and, contrary 'to the guidelines, does not manually actuate the llPI. 5. APPLICABILITY TO DAVIS-BESSE 1 Tho' analysis presented herein required that no auxiliary feedwater is delivered to the SG during the small break transient. This situation is considered highly - unlikely for Davis-Besse 1 because the auxiliary feedtater system is safety grade. liowever, it is expected that the Davis-Besse 1 plant can safely mitigate the

necident'if auxiliary feedwater is restored within 20 r.;nutes.

The analysis presented herein was performed assuming operator action at 20 mirutes to manually actuate the llPI. Due to the low shutoff head of the IIPI pump at Davis-Besse 1, this operator action would not provide adequate protec-tion for this event at Davis-Besse 1. Ilowever, as previously discussed, restsra-tion of the auxiliary feedwater system by 20 minutes is expected to prevent core. 4 uncovery for the lowered-loop plants. Because of the raised-loop arrangement of.the Davis-Desse unit, more of the loop water is availabic to drain into the -vessel. Thus, operator action by 20 minutes to restore auxiliary feedwater at Davis-Besse.dould' provide more margin relative to core uncover than would.cx-1st for the' lowered-loop plants and compliance with 10 CFR 50.46 would be casily assured.' s e 6 ~~ 3 9 4 i 6 1, ,0 _ e I e O e F ,y ' E ~

Item No. 7: (Attachment 1 to Enclosure 1 of Reference 1) 3. Regarding'the presence of noncondensible gases within the reactor coolant . system following a small break LOCA: A. Provide the sources of noncondensible gases in the primary system. B. Discuss the effect of noncondensible gases on: (1) condensation heat transfer, (2) system pressure calculations, and (3)' natural circulation flow. C. Describe any operator actions and/or emergency procedures necessary to preclude introduction of significant quantities of noncondensible gases into the primary system. D.~ Describe operator actions to be taken in the event of a significant accumulation of noncondensible gases in the primary system.

Response

A. Sources-of Noncondensible Gases in the Primary System 1 Table 1 lists the potential sources and amounts of nonconden-sible gases for a 177 fuel assembly plant. However, most of these gases would not be released for small break transients. Appendix K evaluations performed for the 177FA plants demon-strate that cladding temperatures remain low and no cladding rupture nor metal water reaction occur. Thus, these sources can be neglected. Also, the steam generator (SG) is a heat sink only if primary systen pressure is above that which corresponds to the secondary system safety valve setpoint -(% 1050 psia). Therefore, gases present in the core floodina tank can be neglected in addressing the effect of noncondensible on SG condensation. The only sources of noncondensibles which might separate in the RCS are the gases dissolved in the coolant, the: gases in the pressurizer, gases in the makeup and borated water storage tank and gases released from an allowed 1% failed fuel-in the core. B. ~ Ef fects Lof Noncondensible Gases on the Primary System Response Following a Small Break LOCA There are two possible ways in vhich the release of noncondensible gases in the primary system could interfere with the condensation heat transfer processes which occur in the steam generator _during small loss of coolant acci den ts. :If noncondensible gases-filled the U bend at the top of the hot 4 109, then water vapor would have to diffuse through the noncondensible gases before they could be condensed in the steam generator. This would be a very slow process and would effectively _ inhibit natural circulation. Lesser amounts of noncondensibles would reduce the heat transfer by condansation because the vapo,r would have to diffuse through the noncondensibles to get ato the condensate-on the tubes. . Q - - &--e - ~~--- e.------~~ : -~~ ------~ ~~ A: ~.- ~:-- ~ : - - ~ ~ ~ ~ v ~~~ " ~ *~"'"- ~ ~~'

cAsfdiscussedLin. response!to Part.F of this question, the- 'only sources of noncendensibles which might' separate in the 'li.S are the gases . dissolved in'. the coolant, the gases in the pressurizer, gases in Um makeup and horated water storage tank and gases' released from an allowed l'J failed fuel ._in the core. Thus, the maximum-amount of noncondensible gases in the system, assuming all gas comes out of solutien, no noncondensibles are lost through -the breat flow, that there was one percent failed fuel,- and the injection of '6.4 x.10 lbm from the makeup tank and BUST (typical of % 1500 sec of IIPI), 8 'would bei i Dissolved in coolant 563 scf In pressurizer 166 Fission gas 2 Fuel rod fill gas 11 MU tank 24 BWST 14 Total 780 scf 3 This ga would occupy 'a volume of 22.4 ft at a pressure of 1050 psia, the lowest pressure condition in the primary system for which condensation heat removal will' occur. It should be noted that the assumed integrated injection flow does not have a significant effect on the total volume of noncondensibles which might be present in the p'rimary system. Since the volume required tc 3 s completely fill the U-bend in the hot leg is 125 ft, the noncondensible gases will not impede the flow of vapor to the steam generator. The heat transfer during condensation is made up o[the sensibic heat trans-ferred throt-h the diffusion layer and the latent heat released due to condon-sation of thg vapor reaching the interface (see Figure' 1). The model of Colburn and llougenll) gives the following equation for the heat transfer in the vapor phase: (1) Tg ) + Kg Mg'hr (Pg - Pg ) 4 = hg(Tg j g g j g where 2 4 = condensation heat flux, btu /hr ft 'hg = heat' transfer coefficient for vapor layer, Btu /hr f t2 op Tg = bulk ~ temperature, UF g LTgj temperature of interface, F = Kg = mass transfer coefficient, "h I Mg =' molecular weight,.lbm/lb mole hrg = latent heat of vaporization, Btu /lbm -lb P9o = partial pressure of vapor at bulk conditions, Pgi = partial pressure of vapor 'at the interface, lb f x. ft2

• 8 J

"9 g,_ p 4 b .4 i,g g ,,,;;..[ .,,g gm. $[ ^ . j[* 2,,,,,*Mb,,,*.:;4*;$$$$gYhyQ.dGSMT{h M"*7f. Y 4 A4

<.373' r 3 1- 'g 2 p [g _ 3) I E/E * . Kg T pD 2 D = diffusion coefficient, f t /hr 1 L z = height .' lb ft f R = gas constant,1545 lb mole "li T = absol'ute ten.perature at bulk conditions, DR 2 g = acceleration of gravity, ft/hr 3 p = density, lbm/f t 3-po = density'at: bulk conditions, lbm/ft 3 pi:= density at interface conditions, lbm/ft p = viscosity, lb'm/hr ft pam = pai - gao P^' in pao pai = partial pressure of gas at interface, h2 pao = partial pressure of gas at bulk conditions, g j for the application to OTSG condensing heat transfer during small break tran-sients, the term hg(Tgo - Tgi) can conservatively be neglected since the vapor -velocities v/ould be very low.

Thus, 4 = Kg Mg hfg(Pgo - Pgi).

(2) The heat transfer with noncondensible gases present is obtained by iteration. . An interface temperature Tgi is assumed, which fixes-Pgi, and the heat transfer across the liquid condensate film is~ computed from 4 = h (Tgj - T ) (3) f w where 3' 1/4 pf(pf - p )g hfg g k g hp =.943 .pf 2(.f9j - I ) w 3 pf = density-of fluid, lbm/f t ~ 3 p = densi ty of vapor.. lbm/ f t g kg = thermal conductivity of fluid, Btu /hr ft F ~T =. wall. tempera ture, F g . l i ' 59-- m - r r s-r+ ,m r e n- ,,,,,,+,,e --+-w w .-~e,--& v The partial pressure of the gas at the bulk conditions can be calculated from the nule fraction of noncondensible gases. When the heat flux computed from equation 2 matches that computed by equation 3, the proper interface tem;)erature has been found. The impact of noncondensibles on the condensation heat transfer process during a small break was examined for the 0.04,ft2 and 0.01 f t2 cold leg breaks analy-zed for the 177-FA plants. The breaks utilize the SG for heat removal for a significant portion of the transient. Hand calculations were performed, using the theory presented above, to ascertain the effect of non-condensibles on the transient. The amount of noncondensible gases, assuming that all gases come out of solu-tion, would be 2.61 moles. The effect of these gases is to raise the pressure . and primary temperature to obtain the same heat transfer. Assuming that the noncondensibles accumulated only within the steam generator upper plenums and the steam generator tubes, the system pressure increase, due to noncondensibles, 2 break. would only be 25 osi, for. a 0.04 ft2 break, and 40 psi, for a 0.01 ft It should be noted that this effect is predominantly due to the inclusion of2 the partial pressure of the noncondensibles, which is 24 psi for the 0.04 f t 2 break, in the total system pressure. These " break and 34 psi for the 0.01 ft calculations represent the maximum impact as they were computed at the time of maximum condensation heat flux for the respective cases. As shown, the influence of noncondensibles does not significantly effect the condensation heat transfer process. The estimates made are ccnservative in that they assumed all the gas is located in the steam generctors (none is in the top of the reactor vessel or pressurizer) and no gases escape through the break. Thus, the presence of noncondensibic gases in the system should not significantly affect the small break transient. Ca Actions to Preclude Introduction of lioncondensible G'ases into the ~ Primary System Introduction of significant quantities of noncondensible gases into the primary system following a small break LOCA is prevented if the core is not uncovered during a small break. The small break guidelines which have been developed by D&M, are designed to prevent core uncovery by assuring continued ECC injection.

Thus, the amount of noncondensibles which might separate in the RCS is small and would not significantly effect the small break transient (See Part B above).

D. Operator Actions'During Accumulation of lioncondensible Gases in the Primary System A significant accumulation of noncondensible gases vithin the primary system during a small break is not expected. This position is confirmed by small break transient predictions, using conservative Appendix K assumptions, which show that little core uncovery occurs. (2,3,4) As a result of the small core uncovery fuel rod temperature excursions are limited to 1100F; and, fuel red failure or H 2 gas formation due to met,al water reaction will not occur..

Small" amounts of noncondensible gases can be released into the primary system during a small break. For the break size range where noncondensibic gases could have a detrimental effect + (i.e., breaks where-natural circu11a tion is required for energy removal) the' quantities of gases that are predicted to exist within the primary system-are not~significant. For larger quantities-of noncondensibic gases to exist, a core transient that is not predicted must occur. The probability for such an occurrence is believed to be small because of the detailed ~ emergency procedures for post-LOCA condi,tions that have been developed and the extensive operator training that has been conducted in their use. Emergency procedures have been developed to accommodate nonconden-sible gases, to maintain plant control, and to achieve a stable long term cooling condition. Provided below is a brief summary of plant control measures contained.in n,n'esent emergency procedures which.will counteract the effects of noncondensible gases and additional guidance for operator action developed for an inadequate core cooling condition, which will be incorporated into emergency procedures in-the r. cur future. To upgrade the RCS venting and/or degassing capabilities, remote operated hot leg vents will be designed and installed by 1981. Small break-emergency procedures will also be revised at that time to include use of the hot leg high point vents to aid the re-establish-ment of natural' circulation and to vent noncondensible gases which may evolve during small break transient. l I ' CURRENT PROCEDURAL ACTIONS During a small break,_the principle effect of noncondensible gases is to minimize the performance of the steam generators during natural circulation (either sinole phase water flow or reflux boiling). Table 2 lists the primary symptoms and the corresponding operator actions identified in current emergency procedures. As . indicated in Table 2, a restart of the RC pumps (one per leop) is the optimum. action. A return to forced circulation will aid in condensation of existing steam and removal of noncondensible gas (if.present) within the hot leg piping. Noncondensible gases, originally within the loop piping, would then tend to be suspended within the coolant stream and collect within the upper regions of the reactor vessel (RV). A substantial quantity (s 1000 ft.3) of gas can:be accommodated within.the upper region of the RV; therefore, there is good-assurance that natural circulation can be maintained if RC pump operation must be terminated. If the RC pumps cannot be started and/or no' secondary side heat sink is available, the~ operator will utilize the FORV and I!PI 'for core cooling and.RC pressure control until the RC pumps can be restarted and/or. secondary cooling is re-established. s O'

( The above actions are suf ficient to enabic the operator to bring the-unit.to-a stable, long term cooling condition based on expected . plant performance using Appendix K cvaluation methods..Although a large accumulation.of noncondensible gases is not expected under these assumptions, the above actions are believed to be sufficient if the anticipated amounts of non-condensible gases are increased by an order of magnitude because of the larae volume availabic for gases in the upper head of the ;RV and the loss of noncondensible gases out the break. Once stabic long term cooling conditions are established, RCS venting and/or degassing proc.edures can be ' initiated. If the RC pump (s) are operative and pressurizer spray is available, the reactor coolant can be degassed within the pressurizer where the steam-gas space can be vented to the Quench Tank inside containment. If letdown is available, the reactor coolant can -also be degassed utilizing the makeup tank. The reduction of the amount of gases dissolved in the RC will encourage remaining gas pockets within the RCS to redissolve in the water. The operator can monitor the progress of degassinq activities via analysis of pressurizer fluid and/or letdown water samin es. S!4ALL BREAK - If1 ADEQUATE CORE C00LIltG C0flDITIO!1S l An inadequate core cooling condition is not expected for B&W 4 177 FA plants. However, guidelines which identify the symptoms and operator actions for several circumstances, including 6 small ~ break, have been prepared by B&W. This information is discussed in detail in Reference 3. The operator actions discussed in Reference 5 are aimed at restoration of core cooling (restart an RC pump) followed by an . increased rate of plant cooldown and depressurization- (via SG cooling and p0RV operation) to acquire use of the high volumetric flow; capability of~the CFT and Lpl system to maintain core cooling. From a noncondensible gas standptimi, the actions accompiished the following: 'l. Prevention: By initiating corrective action when cladding temperatures are below those for which metal water reaction in significant, gas accumulation is minimized. RC pump E operation (if possible) to restore core cooling and to increase the plants cooldown/depressurization capability is the ~ preferred action. 2. Venting: ' For the core to^be inadequately cooled, the RCS must be in a highly void condition. Therefore, p0RV operation in combination with the break should provide a vent mechanism for the' noncondensible gas that do. exist. 'As discussed in therprevious section, normal venting and degassing- . procedures can'also be undertaken once stable long term cooling is -established. i %.m!MUFME@ ac.VCEEL - es M s=e & :im % ba n E m a m m m ~ M 2 .. ~.

RE F E RE!!C E S 1. Colhurn, A. P. and flougen, D. A., " Design of Coole ~ Condensers for l'ixtures of Vapors with lioncondensing Gases", nd. Eno. Chem. 26(11), 1934 2. R. C. Jones, J. R. Biller, and B.2 11. Dunn, "ECCS Analysis of B&W's 177 FA Lowered Loop flSS", BAW-10103, Rev. 3, Babcock & Wilcox, July,1977 3. l'etter from J. H. Taylor, B&W, to S. A. Varga, ilRC, July 18, 1978 4. ' Evaluation of Transient Behavior and small Reactor Coolant System Breaks in the 177 Fuel Assembly Plant", Babcock & Wilcox, 11ay 7,1979 5. B&U Owners Group Submittal on Inadequate Core Cooling. (Scheduled for submittal i n early llovember) D 4. g l A JE Q-* _g,

• =...

-= r m- --*-) p

.~ 4 h. .e .t v, ' t7 ff Q, TABLE 1 ~f f I

[ 'j S U cr5 07 NCNCONDENSIBl.ES - 177 FA FIXIT-f Total available 13 Mml.-MM WCTIM 1% FAlt.ED FUEL i,

Individaal Individual Individual Individual CO' Total cas Total Gas Total . Total Cas ' Total Total Cas l'. U " Volu=a Volunea mss Masses Vol. Ind. Cas Mass b ases Vol. Ind. Cas msa N sses P 1, source h sef scf Ib. Ib. sef Vol. scf Ib. Ib. sef_ Vol. sef _1b. Ib. +- t 563 H ~ 305 14 N2 ~ I*7 Dissolved in reactor - H2'N2 2 'f ' *I*fE N2 - 158 N2 - 12.3 15 H - 65 5.9 H - 0.4 Pressurizar steam JHy&N2 2 2 e y y. 8F'C'- 3 ~ II N - 5.5 ' 2 2 p cf - . Fressurizer water Hy&N2 30, H - 20 0.91 H - 0.11 2 y space .]. N2 - 10 N - 0.8 2 . Fission gases in' "r & Xa IE6 Kr - 20 65.5 Kr - 4.8 1.9 Kr - 0.2 0.66 Kr - 0.05 1.9 core Xe - 166 Xe - 60.7 Xe - 1.7 Xe - 0.61

y Fuct rod fill gas He & some 1133 He - 1092 14.8 He - 11.5 11.3 He - 10.9 0.16, He - 0.12 11.3 p i N2 & 02 N2 - 32 N2 - 2. 8 N2 - 0.3 N2 - 0.03 02-9 U2 - 0.8 02 - 0.1 02 - 0.01 h.

~ (100~) h Metal water reaction H 416.500 2320 4165 23.2 2 {i MU tank gas spate U &N 726 H2 - 421 26.1 H2 - 2. 3 2 2 N2 - 305 N2 23.8 h. s m 24 H2 - 16 0.71 H2 - 0.09 j MU tak water spacu Hy&N2 %[h 3WST Air (N2 1383 N2 - 902 121.2 N2 - 70.3 8 NZ-8 N2 - 0.62 i d. &0) 02 - 461 02 - 50.9 2 d ; o_ . CF tank gas space N 26,248 2047 2 (two tanks) 4 [.. hIi CF tank water space N 964 75 2 p y. (two tanks} ,~6%J,. { l Assumo:fons - u .t 3 water & 20 std. cc N /Kg water, with water volume = 10,690 f t at 583F and 2200 psia. Jr -j 1. RCS containe 40 std. cc H /K 2 2 g Fressurizer water contains 40 std. ec H /Kg water & 20 std. cc N2/Kg water with Henry's Law relation between water space and steam space at 650F. y.j 2. 2 .eM Vate r volu::io = 825 f t3 and steam volum = 716 ft3 J:Y.[s 3. 72 ilon gases based on inventory in core at 292 EFFD. J ;j-se ' 4. Fual rod gas based on each rod containing 0.0375. gmol He. 0.0011 cmol N2 and 0.00029 teol/0. 2 pN 5. Ntal-water reaction based on 52,000 lb. Zr cladding. 3 J' 0 6. MU tank values based on tank containing 200'f t3 gas space and 400 ft water space at 120F with the trater contalcing 40 std. cc H2/Kg and 20 std. cc N /Kg i.ith Henry's Law relationship between gases in water and in gas spacc. r.i l 2 r a [cQ 7. nW3T contains 450.000 gallons of water saturated with air, i.e.,15 std. cc N /Kg and 8 std. cc 0 /K8' 2 2 h') 3 3

8. ' Each CF t'ar.k contains 1040 fc water and 370 ft gan space with 600 psig N2 at 120F with Henry's 1.aw relation between water and gas.

9. Valxs for 1 failed fuel based on Xe and Kr fission product inventory and fuel rod fill gas (ne) in 1% of fuel rode being released to coolant. '7 10. Values for 1 mtsi-water reaction based on gases in Item 9 above and U2 released from 1% of Zr cladding (520 lb.) reacting with coolant. gs - ,, g. Q. ;i i

b

m s TABLE 2: Symptoms ind Coriective Actions for a Loss of Natural r-Circulaticn During a Small Break (Current Procedures) ' SYMPTOMS _ 1. Saturated coolant conditions 2. Increasing primary system pressdre and t(mperature with stable or decreasing secondary pressure. CORRECTIVE ACTION 1. Maximize llPI (control llPI if a subcooled margin is re-established) 2. Ensure. secondary cooling (i.e., auxiliary feedwater available with proper steam generator level control) 3. Restore RCP flow (one per loop) when possible per the instructions l below. If RC pumps cannot be operated and pressure is increasing i 90 to Step 3.5. 3.1 If pressure'is increasing, starting a pump is permissibic at RC pressure greater than 1600 psig. 3.2 If reactor coolant system pressure exceeds steam generator secondary pressure by 600 psig or more " bump" one reactor coolant pump for a period of approx.mately 10 seconds P (preferably in operable steam generttor loop). Allew reactor ' coolant system pressure to stabilize. Continue cooldown. If. reactor coolant system pressure again exceeds secondary pressure by 600 psi, wait at least 15 minutes, and repeat the pump " bump". Bump alternate pumps so that no pump is bumped more than once in an hour. This may be repeated, with an interval of 15 minutes, up to 5 times. After the fifth " bump", allow the reactor coolant pump to continue in operation. 3.'3 If' pressure has stabilized for greater than one hour, secondary-pressure is less than 100 psig and primary pressure is greater than 250 psig, bump a pump, wait 30 minutes, and. start an alternate pump. 3.4 If forced flow is established,- continue plant cooldown at 100F/ hr.-to. achieve long term cooling.with the LPI/DHR systems. L 3.5-If.a reactor coolant pump cannot be ' operated and reactor coolant j ~ . system pressure ~ reaches 2300 psig, open pressurizer PORV to reduce reactor coolant system pressure. Reclose PORV when RCS -pressure falls tn 100 psi above the secondary pressure. Repeat if necessary. ~If PORV is not operable, pressurizer safety valves will reliev overpressure. m. w Y->YL %l en%. ;~.d nJ ; & %

• w e d 0,,

.l, 3, Q,**;,'; Q M RT:QLl

a O 3.6 liaintain RC pressure as indicated in 3.5 if pressure increases. Maintain this cooling mode until an RC oump is started or steam generator cooling i s established. 3.7 If SG cooling is established:, initiate plant cooldown at 100F/hr. to achieve long terp cooling with the LPI/DilR

systems, s

k. .] e e ..q. e 1 e u. G SP

FIGUFiE 1 e Q ot--- OlFFUSl0!1 LAYER ~ ve I e / / l I>vj l pro YAPOR AllD / Pa' D j ,il0!l C0:10EllSIBLES / - O Ri -Tg9 / / / - JT' g / /o I ~ /,w / / C0!10E!1SnTE .m / P = TOTAL PflESSbRE 2 Paj PARTIAL PRESSURE GF GAS AT lilTERFACE, Ibg/l1 = 2 PARTIAL PRESSURE OF GAS AT BULK 00llDITlut!S, lug /ft Pa = o 2 Pg; PARTIAL PRESSURE OF VAPOR AT litTERFACE, IDg/f1 = 2 Pgo PARTIAL PRESSURE OF VAPOR AT BULK CONDITl0llS, Ibg/ft = T, WALL TE!.'PERATURE, F = TEMPERATURE AT litTERFACE, F Tgj = Tgo_ = BULK TEl.!PERATURE, F REFERfliCE: 1) COLDURN, A.P. A!!D l10VGEll, 0.A., "DESIGil 0F COOLER C0l10EllSERS FOR MIXTURES Of VAPORS NITil NONCONDENSING GASES", Ill0. ErlG. Cl!EM. 2G(11), 1934 -67

• Item No. 7:

(Attachment 1 to Enclosure 1 of Reference 1) 4. Provide a CRAFT-2 simulation fob the first three hours of the TMI-2 acci-dent. The first 20 minutes of this analysis was provided in the "Evalu-ation of Transient Behavior and Small Reactor Coolant System Breaks in the 177 Fue1~ Assembly Plant" (May.7, 1979). We require that the analysis be extended for a period of three hours in order to evaluate the ability of the CRAFT-2 code to evaluate the sequential reactor coolant pump trips and the subsequent period in which natural circulation was lost in the primary system. The analysis should include at least curves for the following parameters: pressure, temperature, void fraction, and flow in the reactor coolant loops.

Response

See Attached. 6 , j

,s 6 CRAFT 2 SIMULATION OF Tile MARCII 28, 1978 THI-2 TRANSIENT I. INTRODUCTION '.In the'May 7,.1979 " Blue Book" reports, a CRNJT2 simulation of the first hour of..the TMI-2 transient was presented. That analysis has since been modified and updated to include more recent esnimates of the net makeup to the RCS during the event. This report presents the results of the latest B&W CRAFT 2 simulation of the TMI-2 cvent and covers approximately the first 2 hours and 20 minutes of the transient. i' The small~ break ECCS evaluation model, which is described in topical report 3 BAU-10104 and the July 18, 1978 letter report, was used,with some "best estimate modifications, for the simulation. Actual TMI-2 data were d combined with availabic information about the operator actions to determine estimates of the HPI and AFW injection times and flow rates. The simulator results (described in detail in the "Results" section) show all the trends and very good comparisons to the actual plant data of system pressure, temperature and pressurizer level. The analysis also ' predicts the time for the start of core uncovery which is in reasonable 4 agreement with the NSAC-1 report. Thus, the CRAFT 2 code is shown to i benchsark very 'well'versus the TMI-2 data and is suitabic for the performances .of small break evaluations. J, METHOD OF ANALYSIS 5- + The CRAFT 2 code which is documented in topical report BAU-10092, was used to simulate the TMI-2 reactor coolant system hydrodynamics. The model uses one node for the reactor building, two nodes for the secondary system, and*23 nodes to simulate the reactor coolant system, including four nodes for the pressurizer. A schematic diagram ef the model is shown in Figure 1. The analytical model used for this sinni,: a 16 basically the same as B&W's ECCS cvaluation model. ' ;Iowever, cu ; < 1a - At assumptions which differ from the' evaluation'model approach, were made in ordet to obtain a "best estimate" . simulation. These assumptions are deceribed below: The initial core power level used.in the model was 102% of 2772. n a. However,Jfollowing reactor trip, the fission produir de' cay heat van adjusted to 98% power operation..The decay heat curve utillecd e .e

1 is 100%, instead of[120% required by Appendix K to 10 CFR 50, -ofit'c INS 5.1 decay'hcat curve. b.. A loss of.the main feeduater purps, which is the initiating transient, was assumed at time zero. In order to account for potential draining of secondary side fluid from the steam ~ generator downcomer into the tube region, a main feedwater coast-down of.10 seconds was utilized. ac.' A turbine trip coincident uith the loss of main feedwater was assumed. This results in the steam generator pressure being controlled by a combination of the turbine bypass valves, the For atmospheric dump valves and the main steam safety valves. the first 90 minutes, the turbine bypass valves control the secoEdary side pressure. In the simulation, these valves were set at.1025 psig.. d. The CRAFT 2. input was adjusted to open the pilot-operated relief valve (PORV) at 8 seconds. This, opening time had to-be input, and the open valve simulated, since CRAFT 2 code does not have models for the pressure relief systems of the RCS. Preliminary TMI-2 data uns used to determine the.FORV opening time. Present TMI-2 scenarios indicate that the P'ORV actually' opened at 3 seconds. As will be shown in the results section, if the CRAFT 2 code had ' ~ an -explicit PORV model, it 'would have predictcd the ' opening at 3 seconds. The'f6' actor scram was chosen to occur at 10 secon' ds lased on c. -preliminary TMI-2 data. Since the CRAFT 2 code does not have _ provisions'for a reactor trip on high pressure, this had to bc . simulated baced on time. f. The.lcak arca utilized for the PORV is l.05 in. and represents ~ .the orifice arca of the valve. The Moody critical discharge correlation was utilized to predict-the fluid lost through the PORV. ~ For the first 4?f minutes of.the simulation, a discharge coefficient (C )i f 0.8 was used. For the remainder of the evaluation a C D D 1.0 was employed. ~ 4 4 3 hV ,- + --n-r

• ,y-

- - = 1^

g. - Actuation of. the liigh Pressure Injection System (IIPI) vas be. ed on ESFA3 signal'of11615 psia. This resulted in the actuation of the 2 llPI systems at 1 minutc and 45 seccnds into the transient,

~ as opposed to'2 minutes and 2 seconds which was the make-up flow J. initiation time at TMI-2. Between 275 and 6100 seconds, the IIPI flow was assumed to be throttled by the operator to an average ' flow of c5 34 gpa. This value is based on preliminary assessment on the not makeup flow to the RCS. No explicit modeling of l'tdown was used, only. net flow was simulated. After 6100 seconds, e an average net makeup (HPI) of 42 gpm was utilized. h. A four-node pressurizermodel was used in the evaluation in order to reduce instantaneous artificial condensation in the pressurizer. This phenomenon, which occurs when the subcooled reactor coolant fluid mixes with two-phase pressurizer fluid, results from the-1 equilibrium model limitations of the code. This model is necessary only to predict the response of the RCS during the initial phase of the loss of' main feedwater event. Also, the pressurizer 4 surge ;line. resistance was updated to reficct more realistically the TMI-2 surge line, l i. Steam Cencrator Modeling - The steam generator model'was modified to account for.the'following phenomena: 1. The'-overall heat transfer coefficient (b'etween primary and secondary) was assumed to. ramp to zero in one minute to account for the delayed auxiliary feedwater injection. i :

12. - Full itcat-transfer coef ficient was reinstatejl at 500 seconds to account for the auxiliary feedwater injection af ter 8 minutes.

3. Auxiliary fcedwater was initiated at 500 seconds with half of thefdesign AFW flow capacity and with the SG level controlled .to 3 feet.- With the reactor coolant pumps on, AFW is controlled ~ by thc.ICS to 3 fect. p o ~4. ' Steam concrator B.was assumed to be isolated at 1 hour and 41 minutes based upon preliminary TMI-2 data. This was simulated ' by.; set ting sthe.hcat transfer coefficient across the B steam a ~ generator:to zero. ~ O 9 e r. ' - 9:

• E.

r

b The auxiliary feedwater control level was manually-raised 5. t to 50% on.the operating range at i hour' and 45 minutes into the transient due to the loss of the RC pumps. 6. Thc main steam safety valves were modeled to open at 5403 seconds, and the feedwater flow was increased at 6100 seconds. This was donc to simulate the steam generator A depicssurization following the operator's attempt tb increase

fceduator flow to steam generator A at about I hour and 34 minutes.

J. The RC punps.in the B loop were tripped at 4400 seconds; the A loop RC pumps were tripped at 6060 seconds. These values are consistent 'with the TMI-2 data. Table 1 provides a comparison of the assumed times for various system actuations and operator actions.to the NSAC scenario. As shoun, the values utilized are l performance during the TMI-2 transient. h t i reasonabic compared to t e ac ua 1 5 e e e t 9 e 9 8 - e 9 e 4 ) -6 4 ~ d 4 +

1 f r n.. '3. RESULTS 3.1' System Prensure Figures 2 and-3 compare the reactor coolant system pressure calculated by CRAFT 2 to the TMI-2 data. Following the-loss of main fcaduater, the pressure in the RCS-rose sharply due to the decreased heat removal across the SG. As shown by Figure 2, the CRAFT 2 prediction overpredicts the pressure during this phase of the accident due to the delayed opening of the PORV 3 seconds in the transient iversus S seconds for the CRAFT 2 simulation, and the delayed reactor trip, 8 seconds for the transient versus 10 seconds for the simulation. If the CRAFT code had. an explicit modelfor the PORV, opening of the valve would have been consistent with the data and a bettcr comparison would have been obtained. After the re-actor. tripped, the RC pressure decreased. The calculated pressure drops below the actual data after 20 seconds. This is apparently caused by the 10 second main feedwater coastdown employed in the simulation overpredicting the drainage of secondary downcomer finid to the SG. Af ter the SG dries out, approximately one minute, the difference between the prediction and the data decreases. Approximately 5 minutes into the transient, the fluid in the hot leg flashed due to the depressurization of the RCS and the system pressure increased. As indicated on Figure 3, the CRAFT 2 code properly predicts the system repressuri-zation time, but overpredicts the actual pressure. The overprediction.of sys-tem pressure is probably caused by the assumed net makeup to the RCS during, t this time period. -Although the HPI was throttled.to a net makeup of 34 spm during this time interval in the simulation, between 4:58 and 6:58, the NSAC . scenario of events indicate that the letdown flew was in' excess of 160 gpm. Thus, it is quite-probabic that there was a decrease in inventory in the RCS ' due to the high letdown over this time period. l At 8 minutes and 18 seconds, auxiliary feedwater flow was readmitted to the SG and prianary system pressure decreased (Figure 3) to approximately 1100 psig and was maintained at that value up to approximately one hour and' 20 minutes. As shown by Figure 3,;the' CRAFT 2 prediction is greater over this period by ~ about 100 psi. The. coolant pressure was measured in the hot leg during the The ac-accid'ent; the~ predicted system pressure shown is the core pressure. tual predicted hot leg pressure is'about 60 psi' lower than the predicted core prensure.< Also, the pressure ~in the secondary side was held in the CRAFT 2 e a 3: Td'%_.. . w, =. _... -.,s..

simulation'at 1025 psig, while the measured value was 1000 psig, resulting in

'an additional deviation. Thus, the CRAFT 2 predicr. ion reasonably follows the transient behavior over this period when the deviations are considered. It should also be noted that the primary system pressure.during this phase of the transient 1s basically controlled by the SG. The CRAFT 2 prediction did not ~ demonstrate fluctuations in system pressure during this period as the sec-ondary pressure of the SG is assumed to be regulated at 1025 psig. The plant data shows that the secondary side SG pressure was not held constant over this period, but fluctuated. At one hour and 34 minutes, the RCS pressure dropped due to an apparent attempt by the operator to increase feedwater to the A SG. The analysis attempted to simulate the depressurization effect of the increased auxiliary feedwater flow by. opening.the relief valves at 5400 seconds and increasing the auxiliary feed-water flow at 6100 seconds. This modeling technique was utilized as littic information is available on the actual auxiliary feedwater flow delivered to the SG'during this peri,od. As shown by Figure 3, this resulted in an underpredic- .tionoftheprimarysystempressureuntil75g0secondsandanoverprediction for the remainder of the transient analyzed. 3.2 Pressurizer Level A comparison of the CRAFT 2 predicted pressurizer level to the TMI-2 data is y 7 provided in Figure 4. As shoun, there are two pressurizer level p'redictions given in the figure. The first, entiti,cd mixture level - CRAFT, is the calcu-l lated mixture icyc1'within the pressurizer. The second, entitled instrumenta-tion reading - CRAFT, is the calculated liquid level that would be "scen" with- .in the pressurizer level tqps and is directly compar'abic to the TMI-2 data, i . The' initial pressurizer response and comparison to the loss of main feedwater event' (first 4 minutes of' the transient) is not easily discernable in Figure I 4. It was.however,-discussed in the May 7,.1979, report. During this phase of the accident, the pressurizer level responded in a similar manner as the system. pressure. Also, the comparison of the predicted to the actual response of the pressurizer. level is similar. That is, the rise in pressurizer level

durisg the first 10 seconds -is overpredicted and. the pressuriz"r level af ter

rcactor trip is underpredicted. The reasons for this are the same as those

~ discussed previously in section 3.1. D 4 e - e i a e- +- i 7 e + y- -- er y-4 g

(The'significant ' aspect of)this comparison is the predicted mixture level re-aponse to the. predicted instrument rcading response during the transient. As shown by Figurc 4,-the predicted instrument respor.se and the measured response are in good agreement throughout the simulation. However, as shown by the fig-ure',' although :the instrument reading is on scale for portion., of the first 101 minutes of the transient, the actual predicted mixture level af ter 6 minutes is at the top of the prassurizer. 'Thus, a two-phase mixture exited through the valve during this entire period. Af ter 101' minutes, only steam was entering the pressurizer through the surge line (note that the RC pumps have been tripped), and the pressurizer mixture had reached a suf ficient void fraction to allow for phase separation at the top of the mixture and only steam started to flow out. 3.3 System Flow As Figure 5 shows a comparison of the predicted and transient loop flows. This shown, the predi6ted flow rates do not match' vell with. the actual data.~ disagreement is caused by two factors. First,' loop flow was measured by Gentillis tubes, wi$ich.are calibrated based on single phase flow. Their actual performance during two-phase flow is unknown. Secondly,~ performance of the RC pumps with two-phase flow is not well understood. In performing the evaluation, a1two-phase pump degradation multiplier based on the semiscale pump tests was utilized. This multiplier results in a sharp decrease ik pump head nce any - o 4 significant voiding is calculated at the pump inlet. As shown, at' 55 minutes, the loop flow sharply decreased due to,this effect. Although the agreement is not excellent, the pump flow calculation does not appear to have significantly, affected'the simulation. 3.4 Hot and Cold Leg Temperatures Figures 6 and 7 show a comparison of the predicted versus actual _ response of the After 5 minutes hot and ~ cold leg ' temperature ' measurements during the transient. and up to the time the core started touncover, the RCS was saturated, and the fluid temperature comparison has'the same deviations previously discussed in section 3.' 1. ~ Af ter the core. starts to uncover, 'which ' occurs at approximatcly 110 minutes, the ~ indicated superheated steam (Figure 6). How- - hot Icg. temperature measurement This is due to cver,.the CRAFf2Lprediction does not exhibit this behavior. Lthe one-node representation of the core and the equilibrium assumpt' ion of.the 4 e 4 c

• s x

CRAFT code. As long as fluid is predicted to remain within the core node, re-gardless of the actual amount of core uncovery, the one-node representation calculates the exiting steam temperature to be saturated. Ilouever, the actual physical precess results in saturated steam at the top of the core mixture level. This steam superheats as it receives energy from the uncovered portion of the fuel pins. A multinode representation of the core would be necessary to predict the hot leg temperature response during this period. 3.5 System Void Fraction The average system void fraction evolution for the primary system, excluding the pressurizer, is shown on Figure 9. Due to the continued loss of RCS in-ventory through the PORV and the inadequate net makeup to the RCS, tiu: system void fraction increases almost linearly from 10 until 101 minutes into the transient. At 101 minutes all the RC pumps have been tripped. At this time, the RCS liquid inventory is distributed as follows; the RV is filled to slightly above the top of the core; the loop seal in the B loop is full; the A loop has very little invent'ory. During the subsequent. 30 mir.eues, the RV invec. tory is~ .Because boiled-off and the steam 'is condensed by the A loop steam generator. of the lowered loop design, this inventory remains trapped within the A loop pump suction piping and the steam generator. During this period of time, the core becomes uncovered. Thus, since the process is a redistribution 'of water ~ within the RCS uith the only fluid loss being steam vented threugh the PORV, the system average void fraction does not change significantly. 3.6 Core Mixtr e Level The calculated core misture icvel for the transient is given in Figure 8. As How-shoun, no core uncovery was calculated while the RC pumps were operating. the level in the cver, closely follouing the termination of the RC pump flow, core decreased. Core uncovery was calculated to start occurring at 105 minutes into the transient. This comparcs reasonably well with the USAC prediction of approximately 103 minutes. Thus, the calculated loss rate through the PORV and the not makeup to the RCS must be in reasonable agreement with the actual behavior during the TMI-2 incident. However, this As shown by Figure 8, the core was picdicted to totally uncover. result occurs due to the insufficient spatial detail in the core region..The simulation assumes that all core heat is removed and deposited in the fluid. This results in an overprediction of the core boil-off once the core'is'un-A more detailed core model is necessary in order to predict how mich covered. core heat in deposited in the liquid region for subsequent boil ^off of the % I

1 Ilowever, since core liquid and how much energy is used to superheat the steam. the simulation was baulcally made to determine how the core uncovery occurred, the refined core model was deemed unnecessary. l 4. CONCLUSIONS _ As demonstrated, the CRAFT 2 code simulati n predicts reasonably well the system In behavior during the first 2 hours and 20 minutes of the TMI-2 transient. the core uncovery time is predicted within a few minutes of the inferred

fact, core level response given in the NSAC report.

Therefore, it is apparent that the n'et makeup to the RCS was very low (approximat ely 34 gpm) over this period Also, it which resulted in uncovery of the core and subsequent core damage. is shown that the CRAFT 2 code is abic to predict the system hydrodynamics dur-ing a small LOCA and is suitabic for licensing calculations. e e e 9 e 9 e I 4 4 e 9

)

• REFERENCES 1

" Evaluation of Transient Dehavior and Small Reactor Coolant System Breaks in the 177-Fuel Assembly Plant," Eabcodk & Wilcox, May 7, 1979. 2 "B&W ECCS Evaluation Model," BAR-10104, Rev. 3, Babcock & Wilcox, August 1977. 3 Letter J.ll. Taylor (B&W) to S. A. Varga (NRC), July 18, 1978. 4 " Analysis of Three Mile Island, Unit 2 Accident," NSAC-1, July, 1979. 5 R.A. liedrick, J.J. Cudlin, and R.C. Foltz, " CRAFT 2 - FORTRAN Program for Digital Simulation of a Multinode Reactor Plant During Loss-of-Coolant," BAW-10092, Rev. 2, Babcock & Wilcox, April 1975. .O d t G e 9

'y-Table 1. Comparison of CRAFT 2 Assumption to MSAC Scenario l Time, hrs: min: sec Event NSAC CRAFT 2 Loss of feedwater flow / turbine trip 0:0 0:0 l P0RV opens 0:03 0:08 Reactor trip 0:08 0:10 IIPIs actuated 2:02 1:45 11PI throttled 4:38 4:35 Auxiliary feeduater block valves opened 8:18 8:20 Reactor coolant pump 2B stopped 1:13:29 1:13:33 Reactor coolant pump 1B stopped 1:13:42 1:13:33 Steam generator B isolated 1:42:00 1:41:40 SG A level raise'd to 50% on operate range 1:40:00 1:41:40 i Reactor coolant pump 2A stopped 1:40:37 1:41:00 . Reactor coolant pump 1A stopped

• 1:40:45 1:41:00 W

4 4 ,M' e 4 9 e _79_ p

,. ~. - - - - -. 0 a ?., ^ S, 0 g e g "*) "c.' d $ 8. "U e "3 -4 h U a** i Si 35 O"es j 'E ".o b c '$ c eu 4 o w 4

• 3 s ::

2.4 U O ** ~ ~- a;

• 5 u

na n=3 I .n <I g Y g 9c l g y <n [6 ,,3 .3 ,1 of rn l 't .= .= m en - n l l h re ~1 l AA/ ce N SJ e. Co @l;- ari m g N N D rg [-4 N L--- m e,., g-4 H l u t: O r4 tu _ IIN. /'V T_ El [ \\.1 i V ~ n u '.I .g is. D .J y d. m n a n A V L -- O:: I c 1 1 16 ec 'l /N L_IlT/ N I t r4 l o /Tf T O GQ m! w,pt' 3 - V_ _)U2G), .E ( ~ e m. a e e = = @j = C4

• W G

e. 9 t c) l t _. - -380-

,k. 2000 - FIGURE 2 SYSTEl4 PRESSURE VS TIME j 'l 2400 -l }, _ Tla!-2 DATA Ol -.. CRAF T -2 Sl!,1UL ATl0!! [ \\ \\. j 2200 p'( l I I 2000 j t .2 i E- \\ or. \\ t 1800 \\ m 0 \\ <t \\ N .N.N N 1000 .%u fi .1 s m 1400 1200 -1000 I I I I 0 1 2-3 4 5-Tinc, Hinutes 4 ' l ') +-

's ~ FIGURE 3 REACTOR C00LANT PRESSURE 'N Time, f4inutes 20 40 60 80 100 120 i40 300b 6 i i 4 4 4 2500 CRAFT 2 - Sifr!ULATION -- -- TL!! ACCIDENT 2000 ~ n a. P. o, u 4 = - $ 3 1500 A i c_ 1 / \\ o V \\ \\ g N s- - --, m - - A* 4 1000 ~~ k ~% /,/ 500 0 1000 2000 3000 4000 5000 6000 7000 8000 Time, Seconds e s% 9 -.-n

/ _ s J C3 6-3 'l. I 1 g y I i a \\ l (. e4 ~ -i. i 't ' I a t l ce .o \\ .I e3._, .S' 4 tr.a g H ts. c3 taJ gea i o m m ,/ m e Z N. O ~ .B --. CC L*4 H .3 > 4 &3. LAS D-= &3 O J Z t ta U3 taa (C LAJ L1 Ch. En 22 e C3 (C e4, H H e .\\ 6 ><, &3 + y Z M I .a ~ >-= om _J [ -s Ill> \\. H m c3 LA 1 1

3

• E' 18]

"h ,N I O ~ H N s . u Q"5 l ' ;~~ in l U) l a + ti! LC cn Q. 1 T s. 1 ~ It! g): C2 6 ik. + O g 1 u-I i i m 't 1 1 l o g O I i g 1 a t9 .J I i ) T t -f I a l\\ N N

• • • • • --. D ; -.___

--.._,,m w 3;;;,l,*w,.%. I 1 ~....~'..-L-_-: E a. O ' a O c-3 <3 c3 o C3 E.3 C3 o g3 ' q. 3 41) WB 47 Ev3 CN e. ~ ul 'laaal 4 1 0 ~ y

.f / e c3 V 4 ct' CD t.3 H H LJ.J" LaJ =1; 4 J J C3 C3 O O O ::i2 E N mcH F-= e e e asC CD 4 CO A CL A O.

• O O O O O'O O O J J J J

.e o e /.#. l 8 e O l e .l

• s,j.#.

O e e l l ./ = n / l0 i o. = O u. O m o " ::= uJ f-as ......: y p....e.:- .L:- g \\. se~**~~.,_. .,,,%y.... .H )- 1 O / ../ ( 'I y I - O (*. / O g) a-j- -) e .s H u. ! / _ O .I e

)-

I' i

• *. _)

x/ - O ,e n 'k .o e, 9 t 1 f I I' O .O O O O O O O O O O

O

..O O O C3 O O O N CO T N o* s/tq [ 'M0lJ -84 :

.f .e- . /_ f Y, I C3 d %f O

r..

., / - o . a g *% to . / a

• N.g

............................................ %..s.... m. ,_..-..-....--..=--=..===~**==-****'%,. ~~~*""% ! * *.(.S a %'~. g'h e N ~ ~ ~. ./- ! (* ~ o C A. o [g I e r-s } ( ) .\\ / .. ),:

h.. /

l / I .f* c2 ./ ),/. Y.- . c2 C) $~

• *I:

.f*. O. o ~ p,.J' m 9 .I.

k. t

\\ N.NN.*. I N; t .cc =ct e l & W 1 ]4 " c 4 4 . ]. I 9 i ) o c:2 c2 /.- c3 '~ 7 k.'@,< N y.3 .c w ~n '.-) i e

:: r H

~ 9-E c: ',~ r u. u.

• 2

'cr';. Cr:'. "4 i f. ~ C3 0". fL O 3 o Z Z U U T. s O S Q[;# -e I ~

c m to <<

h & =q SO n! .oi. O c, r. o o c) o i... (,g% t

c m

_c ::: . \\,.9 ~) W D

• \\

i ; k \\ \\s )i. i . o a a . l.1.: 1 i o e l /. .o ./ N. { .l k . (N ') { ll . -4 o ~ n; ~r .i l)

I-O t G o

/)

T O .N

• /'

1.j 3. e ** O ..*** g h 'N .p" a r.r.-4 aa ,. *,... *..... y.w.:=. - s/ pf.,/* e* . ** e.t*gs{" 7f yg y 1 I l a c3 c3 c3 C3 - C3 C3 =r N C3 m = a;- m a, .u .a m. 6 e . -,. ~.,

e .t -t .I W.W I 4 at I / o o 1 / N. N.. p 55N ^ N o ~ w w w w i N W L l e O cm -t n

• /

o cc cy: cr: C2: m-M 1% O O / < m m < n cl. n n o o o o o o o o J J k 8' o c3 o ca c3 C3 ,g O J J a J o o o o o o

• .}

', / to O O O O l~ W 9-= H I i 1 \\e i l e g w I l c3 I t o c3 O _q. sn u) O' j\\ l lil \\ n, h 2 uJ i s m e-g m m p

r3 c2 g

a-c: c2 (D o .cn f =s bl o v i c: as p

E v2 l

o e 8 co .\\ e J - a> .E3-g -O' .E- ,- g g J H H g. C3 V~ C3 1, 1 O ~ I w (C l a a O l - v H ] UL. i i-O + $l ~ 'r// N i 1 i f .: 1 7/ o fj N .o / e g .,), / ..J'/ <$,-7 ~~ s s g h.} f I 3 I I o c3 c3 cm c3 c3 o to N, CO "J" C3 CD LO ' (O U3 Lf3 LD v J 'ajillej adual 9 e u. a

FIGURE 8 CORE MIXTURE LEVEL 12-r .1; \\ \\ CRAFT 2 SIMULATION \\. i CORE LEVEL PREDICTION 10 V, + 5 FROM REFERENCE 4 0 t \\ s \\ \\ ~. \\ \\ .. S, 6 .\\.- .\\ Y .c \\ ~ =a g 4 \\' g s, N N % .2 [ _ i i i i i i i 6000 7000 8000 9000 10,000 11,000 J Time, seconds

2 +-.. -., a. m -e m .e-- I. ,'.-f; e g }. n. } d a s - 1; -gg-4 h. t 1 J e \\ t, I 5 4 a k g es PD o 8'* .e E 4 l 1 l o e P4 ee i ( ) c3 M, O t-a . t)} C3 O'* F, o OJ -- La.. ' e H n L ci s8 I. t .( a, g 6.~ 4' w. c2 so - w e as b H - v1 - u 42 gg 3 . c e . (?. o. c => >0 M On 7 D,d a d 40 t= ~ o.n a p b e a9 M-N.. eg l' es O' es > we 1 7. 'O - ta. and. m ,M.- u - M . Y g ~ g j.- . taJ i C.a et a b - m 2 - ut e Y e4 1.- .s m e a,, r3 e j 34 ** ' Ch - -. sg M D= laj - ry-c3 ~ 6 4 Ce g i. + gc3 4 O e ).. 4 Et e-o ?- 't 1 ' s e e f t c3 O . tas - se t. CD r3 7 es es. ~ca. t9- - c3 N =g. c3 eg s "spleA viaisis M - t h h t v. 1 4 f a 4 w .,,y.-,-, .. +.- -y,,, ., - - - - +,,.. ,.,,-,y, 4 y.+ e4 -e 4 y

- Item'No. 7:.(Attachment 1 to Enclosure 1 of Reference 1) 5. Perform an evaluation of the recent Semiscale small break experiment (S-07-10B) with your small break computer program. This request was sent to D. Holt (Chairman, B&W 0wners' Group Subcommittee on TMi-2 Follow-up) from D. Ross on July, 16, 1979. Copies of this letter were sent to all B&W Licensees.

Response

This subject has been discussed in the Babcock & Wilcox letter from James H. Taylor (B&W) to R. P. Denise (NRC), " Analysis Prediction for Semiscale Test S-07-10B," dated October 9, 1979. I.,, I i F r . das.a 2" -'T

• M1?b[-. ~ - 1" ' C.'w

'.-.*;i.'---. .~--e~' b .w

A

-Oi P

r-- Item No. 7: (Attachment 1 to Enclosure 1 of Reference 1) 6. Pretest calculations of the Los's of Fluid Test (LOFT) small break tests shall be performed as means to; verify the analyses performed in support of. small break emergency procedures and in support of an ~ eventual :long-term verification of compliance with Appendix K to 10 CFR Part 50. This item is discussed'in Recommendation No. 2.1.9 of NUREG-0578.

Response

This subject has been discussed in Babcock & Wilcox's letters from James

11. Taylor (B&W) to D. F. Ross, Jr. (NRC), " LOFT-L3-1 Pre-Test Prediction,"

dated November 20, 1979 and December 14, 1979. . 1 ,g. -.,<a

Item No. 8: This item is discussed in Section 8.4.1 of NUREG-0560 and endorsed in the staff's evaluation of each plant. Some of this work has been completed; however, the scope and schedule of this requirement has been superseded by' recommendation 2.1.9 of NUREG-0578. In a meeting with the staff on August 9, 1979, B&W and the B&W Owners Group presented a program by which they intend to satisfy this requirement. Subject to incorporation of the comments given by the staff at the August 9 meeting and additional com-ments discussed with B&W by phone (Z. Rosztoczy (NRC) and E. Kane (B&W)) on August 14, 1979, the staff expects the proposed program and schedule for completing this item will be acceptable. By August 31, 1979, each utility should provide a written program outline and schedule for com-pletion of this item.

Response

GPU is involved in an Abnormal Transient Operating Guidelines (ATOG) Program that is being funded via the B&W Owner's Group. The goal of the ATOG Program is to generate Operator Guidelines. The guidelines will provide the operator the means to determine the appropriate ac-i' tions during transients. The desired goal is to make the guidelines symptom oriented. They will provide specific instrumentation to be monitored, expected plant alarms, and probable plant conditions. Based on a set of plant conditions, a set of recommended actions, directed toward placing the plant in a safe shutdown condition, will be given. In addition, the guidelines will provide expected plant conditions resulting from correct or incorrect operator / equipment responses in the performance of the recommended actions. Included in these recommended actions will-be the time available to the operator for recognition of the plant conditions and for the im-plementation of the recommended actions. Where necessary, the re-quired time to reach a certain plant condition will be specified. If the plant has not reached that condition, then the operator will be directed to another set of recommenr'ed actions which is governed ~ by plant conditions at that time. The Operating Guidelines will consist of two basic parts. Part I will provide detailed operating instructions to the plant operators of spe-cific symptoms to. observe and specific actions to take as a result of these symptoms. Part II will give the bases for the detailed instruc-tions in Part I. The conventional event-oriented procedure format will not be used for Part I of. the guidelines. It is recognized that. guidelines which are . concise and clear are desirable for use in short time frame, high stress situations. B&W intends, therefore, to develop an improved format for Part I of the guidelines. A new vehicle to convey technical decision,

6 . Response to Item No. 8'(Con't.): making informatIon to the operator wii1 he created. - An independent "Iluman l' actors" engineering consultant will be employed to assist in this erfort. Present schedules call for completion of the Operator Guidelines by October 1,_1980. , c.

~ Item No; 1: - (Small' Break LOCA - Reference 2) f Transitions' from solid natural circulation to reflux boiling and back to solid natural circulatioin may cause slug flow in the hot leg piping. 3y Luse of analysis land/or experiment address.the mechanical cffects of the induced slug flow on steam generator. tubes. 1 RESP 0 HSE. The' loads imposed on -th6 tubes of the OTSG during the postulated " slug flow" have been conservatively evaluated and found to be acceptable. Based on' very conservative assumptions, the end loading on each tube will be 21.5 lbf compared to-a theoretical buckling load of about 700 lbr. It was assumed for this analysis that a water level has been established in - the hot leg piping and inside the tubes of OTSG. ' The transient. consists of a " front" of solid water impinging on the primary face of the upper tubesheet.- The flow was assumed to be equal-to full 100% power flow (about 70,000,000 lb/hr). The load is assumed to be a scddenly - applied load. The upper tubesheet is-conservatively assumed to offer no resistance to the load and the lower tubesheet is assumed to be fixed so that the entire load is absorbed by tubes directly under the primary inlet nozzle. The flow is assumed to not follow the diffuser so that the velocity impirging on the tubesheet is the ~same as the velocity in the 36-inch nozzle. Hot leg temperature is assumed to be 6050F The velocity in the 36-inch pipe would be 64.4 f t/sec. By use of the momentum ' equation, the steady-state force on the upper tubesheet due to the ' velocity ~would be 16,080 lbr. Assuming a suddenly-applied load, the momentary force would be 32,160 lb. There are about 1500 tubes in a f 36-inch diameter circle. Thus the 32,160 lbf will _ result in 21.5 lbf per tube. Since the cross-sectional area of each tube is 0.070 in, the . momentary axial compressive stress in these tubes wo'uld be 307 ' psi. 4 D A i 9 f [ e t 6 ..l

Item No. 2: (Reference 2) llow are the HPSI pumps protected against deadheading if the system re-pressurizes? If operator action is required, show that there is suffi-cient time for operator response. How is this covered in guidelines for emergency procedures? (Applicable to Davis-Besse 1 only)

Response

As indicated above, this question is only applicable to Davis-Besse 1, and thus no response is required for TMI-1. 1 Item No. 3: (Reference 2) Evaluate the impact of RCP seal! damage and leakage due to loss of seal cooling on loss of offsite power. How long can the RC 1eals sustain loss of cooling without damage?

Response

The reactor coolant system pumps employ a controlled leakage seal as-sembly to restrict leakage along the pump shaft, as well as a secondary seal which directs the controlled leakage out of the pump, and a vapor seal which minimizes the leakage of vapor from the pump into the con-tainment atmosphere. A portion of the high pressure water flow from the =akeup pumps is in-jected into the reactor coolant pump between the impeller and the con-trolled leakage seal. Part of the flow enters the Reactor Coolant Sys-tem through a labyrinth seal in the lower pump shaft to serve as a buffer to keep reactor coolant from entering the upper portion of the pump. The remainder of the injection water flows along the drive shaft, through the controlled leakage seal, and finally out of the pump. A small amount which leaks through the secondary seal is also collected and removed from the pump. Intermediate cooling water is supplied to the thermal barrier cooling coil. In the event that offsite power is lost, intermediate cooling water would not be supplied to the thermal barrier cooling coil. Seal injec-tion water would be temporarily interrupted,.but would be reestablished upon loading of the makeup pumps on the energized emergency buses (Auto Load Block 1) and starting of the pumps. The reactor coolant pump seals will retain their integrity indefinitely via cooling supplied by only seal injecti;n water. Even though a complete less of seal cooling is very unlikely, as dis-cussed above, an evaluation of the consequences has been performed using engineering judgement and the limited experience applicable. The evaluation shows that leakage would not increase appreciably for approximately 10 minutes and would not be severe for up to 60 minutes. In this evaluation, it was assumed at time O that the pumps are stopped when both seal injection and seal cooling are lost, the seal flow return flow valve is open and initial leakage is at a normal maximum of 2 GPM for mechanical face type seals. (Note that pumps with a first stage film riding - hydrostatic seal may leak up to 5 GPM, but due to the large internal heat sink of this type of seal, the projected times in this evaluation will be about the same'.) The capability exists to nanually reinitiate seal injection flow and to close the seal injection return flow valve without offsite power and to restore cooling water flow when offsite power is availabic. The seal cavity temperatures and seal leak rates for the first 4 to 5 minutes from time zero, will remain essentially stable due to the mass,

Response to Item No. 3 (Cont'd.): 1 of the~ heat sink at'the shaft, seal cartridge and pump heat exchanger. This time period could be extended by about 2-3 minutes if the seal re-turn valve is closed within 90' seconds. With the seal return valve open, when the temperature in the seal cavity star.ts to rise, it will increase at a rapid rate. The seals will be passing steam in an additional 4 or 5 minutes. If seal injection can be gradually reinitiated or if the component cooling water flow is started within about 10 minutes the temperature ramp will be turned around, and although some internal damage may have occurred, the seal system will gradually stabilize and return to approximately the initial leakage rate. Closure of the secl return valve within this time frame is most effective in slowing the rate of temperature rise on those pumps that had low seal leakage at time zero and have not reached the point of rapid temperature increase. Closure of the seal return valve shortly af ter time zero would have reduced the heatup rate by as much as 75% for low leaking seals or 50% for high leaking seals. If cooling continues to remain unavailable, the seal cavity temperature will continue to increase and is predicted to reach at least 350*F in 20 minutes. At this time, the shaf t directly above the seals will be about 300*F and the heat exchanger will be at full system temperature ( 540*F). The rapid restoration of cold seal injection water af ter the seals and pump parts are hot will shock all of the hot parts causing distortion and possible cracking of seal parts which could lead to an increased leak rate, llowever, it is felt that this will not cause an appreciable increase in leakage on a static pump. It is preferred that component cooling water be the first source reinstated until the temperature in the seal cavity have returned to normal and have stabilized. If com-ponent cooling water cannot be reinstated,>.then cold seal injection may be initiated, preferably at a gradual rate (approximately increasing the rate at one gal. per minute). After steaming conditions are reached, significant seal degradation would not be expected up to a period of approximately one hour af ter time zero. The clastomers which make up a part of the' seal assembly i will start to soften at approximately 300*F and can start to extrude before reaching 500*F. The amount of extrusion is based upon time, temperature and annular clearances. Experience has shown that leakage of seals because of elastomer extrusion does not increase appreciably within the first 30 minutes. It is estimated that under the worst conditions, leakage on a static pump may reach 5 GPM in 30 minutes and 10 GPM in-60 minutes. After the pump _ experiences high seal cavity temperatures, the following parts must be inspected and replaced prior to ' operation: ;

- - ~. i Response to Item No. 3 (Cont'd.): a. Seal package - replace all elastomers and seal faces, inspect all structural parts including bolts for distortion, cracks, etc. b. Water lubricated carbon bearings - inspect for cracking and steam cutting. c. Perform pump-rotor alignment check. d. Monitor shaft and frame vibration on pump start to determine if thermal shocking has produced a bow in the pump shaft.

- _}}