ML19263A251

Board Notification 84-057:describes Recent Calculations Performed to Investigate Equipment Temp Response in Ice Condenser Containment

ML19263A251
Person / Time
Site:	Catawba
Issue date:	04/02/1984
From:	Novak T Office of Nuclear Reactor Regulation
To:	Foster R, Kelley J, Purdom P Atomic Safety and Licensing Board Panel
Shared Package
ML19263A252	List: ML19263A251 ML20087N716 ML20087N720
References
TASK-AS, TASK-BN84-057, TASK-BN84-57 BN-84-057, BN-84-57, NUDOCS 8403140260
Download: ML19263A251 (4)

Text

Central Files Only April 2, 1984 Docket Nos: 50-413 and 50-414 MEMORANDUM COR: The Atomic Safety and Licensing Board for Catawba Nuclear Station, Units I and 2 (J. Kelley, P. Purdom, R. Foster)

FF0M:

Thomas M. Novak, Assistant Director for Licensing Division of Licensing

SUBJECT:

BOARD NOTIFICATION - EQUIPMENT TEMPERATURE RESPONSE IN AN ICE CONDENSER CONTAINMENT (Board Notification No.84-057)

In accordance with NRC procedures regarding Board Notifications, we are forwarding the following two documents from Sandia National Laboratory (SNL) which may be material and relevant to the matter of temperatures used to qualify equipment located in an ice condenser containment. The appropriate Boards and parties are being informed by copy of this memorandum. The two documents are:

1) Memo to J. Larkins (NRC) from V. Dandini (SNL) and W. McCulloch (SNL), January 24, 1984

2) Sandia Rough Draft Report, "HECTR Analysis of Equipment Temperature Pesponses to Hydrogen Burns In An Ice Condenser Containment," February 1984.

These documents describe recent calculations performed to investigate the temperature response of equipment in the Sequoyah containment during a spectrum of severe accidents. These calculations are intended to simulate the effects of simultaneous hydrogen and steam releases to containment and the effect of hydrogen combustion.

Furthermore, estimates of the prob-ability of various accident sequences are provided in the report.

For numerous accident sequences, the analyses predict a higher equipment temperature than that which was used for equipment qualification (440 K).

Based on a review of the accident sequence probabilities and the associated equipment temperatures, the report concluded that there are several accident sequences, with a probability close to that of the sequence used as the design basis for hydrogen igniter systems, which result in equipment tenpera-tures higher than the qualification temperature. On this ice condenser plant (see Supplement No. 6 to the Sequoyah SER), the staff indicated that equipment survivability was to be evaluated for a spectrum of accidents as a confirmatory item, through the NRC's Hydrogen Burn Survival Program at 314026 zA vff

.p_

Sandia. Thus the SNL documents represent the current status of these con-firmatory calculations.

For Catawba, the SNL documents are potentially relevant to SER Sections 1.S(7) and 6.2.5. and to Unresolved Safety Issue A-48 in SER Appendix C.

At this time it is difficult to assess the accuracy of these calculations and, therefore, the reports' conclusions. There are, from a cursory review, a number of errors and items which require clarification.

The staff is currently reviewing the documents and will advise the Board of results of our evaluation once completed.

A

/ T omas M. Novak, Assistant Director for Licensing Division of Licensing

Enclosure:

As stated cc: SECY (2)

OPE OGC EDO Parties to tPa Proceeding See next page 0

.kD DL:L5 #4 W C[d 4

DL AD:L:DL KJabbour/hmc EIdensam RStarke Novak 3/$/84 3/ ^ 84 3/ /84

$4

.DJti 3/tth7

DISTRIBUTION LIST FOR BOARD NOTIFICATION Catawba Unit 1 Docket No. 50-413/414 ACRS Members John Clewett, Esq.

Dr. Robert C. Axtmann 3

Dr. Richard F. Foster Mr. Myer Bender Robert Guild, Esq.

Dr. Max W. Carbon Dr. Frank F. Hooper Mr. Jesse C. Ebersole James L. Kelley, Esq.

Mr. Harold Etherington Dr. Robert M. Lazo Dr. William Kerr Karen E. Long, Esq.

Dr. Harold W. Lewis Morton B. Margulies, Esq.

Dr. J. Carson Mark J. Michael McGarry,III.Esq.

Mr. William M. Mathis Palmetto Alliance Dr. Dade W. Moeller Spence Perry, Esq.

Dr. David Okrent William L. Porter, Esq.

Dr. Milton S. Plesset Dr. Paul W. Purdom Mr. Jeremian J. Ray Mr. Jesse L. Riley Dr. Paul C. Shewmon Alan S. Rosenthal, Esq.

Dr. Chester P. Siess Howard A. Wilber, Esq.

Mr. David A. Ward Mr. D01ald R. Willard Richard P. Wilson, Esq.

Atomic Safety and Licensing Board Panel Atomic Safety and Licensing Appeal Panel Docketing arid Service Section Document Management Branch e

CATAWBA (For BNs)

Mr. H. B. Tucker, Vice President Nuclear Production Department Duke Power Company 422 South Church Street Charlotte, North Carolina 28242 a

cc: North Carolina MPA-1 P.O. Box 95162 Raleigh, North Carolina 27625 Mr. F. J. Twogood Power Systems Division Westinghouse Electric Corp.

P.O. Box 355 Pittsburgh, Pennsylvania 15230 Mr. J. C. Plunkett, Jr.

NUS Corporation

?536 Countryside Boulevard Clearwater, Florida 33515 Mr. Pierce H. Skinner Route 2, Box 179N York, South Carolina 29745 North Carolina Electric Membership Corp.

3333 North Boulevard P.O. Box 273C6 Raleigh, North Carolina 27611 Saluda River Electric Cooperative, Inc.

207 Sherwood Drive Laurens, South Carolina 29360 Mr. Peter K. VanDoorn Route 2, Box 179N York, South Carolina 29745 James P. O'Reilly, Regionsl Administrator U.S. Nuclear Regulatory Commission, Region II 101 Marietta Street, Suite 3100 Atlanta, Georgia 30303

Sandia National Laboratories AIDucue!Oue New M e s eCC B7165 January 24,~1984 i

Mr. John Larkins Severe Accident Assessment Branch U.

S. Nuclear Regulatory Commission Mail Stop 1130SS Washington, D. C. 20555

Subject:

MARCH-HECTR Analysis of the Effects of Selected Accident Scenarios on Equipment in an Ice Condenser Containment Introduction Camp, et. al.,

in their report, MARCH-HECTR Analvnis of Selected Accidents in an Ice Condenser Containment. describe the pressure / temperature environment, in the Sequoyah containment, resulting from hydrogen burns precipitated by several accident sequences.

The report is currently undergoing review for publication approval.

The HECTR runs described in the report also included temperature calculations for several generic surfaces representative of safety equipment at three locations inside containment.

The temperature calculations and results were not included in Camp's report.

This letter will summarize the results of those temperature calculations.

Description of Surfaces Three surface types were considered in each of three different locations inside containment.

The locations were:

the upper compartment, the upper plenum, and the lower compartment.

The surface types were:

thin valled aluminum (.125 inch), thick walled aluminum (.67 inch) and the.25 inch thick steel cover plate from a Barton pressure transmitter.

Scenarios Four basic accident sequences with variations on each were considered.

These four sequences are given in Table 1.

Y D R.

's'/Wfuyus Q

/3 po f

4 -

.e Mr. John Larkins January 24, 1984 Table 1 HECTR Sequoyah Analysis Basic Accident Sequences Seouence Description SD Small break LOCA (<2" dia) with ECCS failure 2

SD Intermediate break LOCA (2" i dia 1 6") with 1

ECCS failure SH Intermediate break LOCA with failure of low 1

pressure recirculation system TML Transient with failure of power conversion, secondary ste m relief and auxiliary feedwater system.

S H and TML sequences were Several variations of the S D.

1 2

considered along with a few for the S D sequence.

I Results The results of the surface temperature calculations are given on the enclosed work sheets.

Surfaces 4 8,

and 17 are the thin-walled component, surfaces 5.

9, and 18 are the thick-walled component and surfaces 6.

10. and 19 are the pressure transmitter cover plate.

Three conditions were considered:

initial temperature (Ti). peak temperature (T ) and p

temperature increase (AT = Tp - Ti).

The initial temperature is the temperature at the initiation of the first burn.

In cases where no values for Ti and aT are given, no burns occurred in the compartment where the surface is located and T is the result of the LOCA environment and burns in p

other compartments within the containment building.

An explanation of the color coding is given in Table 2.

Table 2 Sequoyah Work Sheet Color Codes Ti 1 440K Green Bordered Box T

1 440K Red Bordered Box SbK$ A T < IOOK Blue Box 100K i a T < 200K Orange Box 6 T 1 200K Pink Box

~'

s Mr. John Larkins January 24, 1984 Particular attention was given to those analyses which calculated a peak temperature of 440K (332*F) or higher._ This temperature is representative of maximum LOCA qualification temperatures for safety equipment, and higher temperatures generally exceed the level to which operability has been a

deubnstrated.

Those cases which resulted in peak temperatures of 440K or greater are described in Table 3.

When known, approximate probabilities of their occurrence per reactor year are also given.

These probabilities refer only to the accident sequence and are independent of such factors as the amount of zirconium reaction and code parameters.

Table 3 Accident Cases Resulting in High Peak Temperatures (with hydrogen ignition limit set at 8% unless specified)

Case Description Probability

• S D Casas (with 75% 2r reaction unless specified) 4.8x10-6 2

S DX, failure of air recirculation fans 8.6xlO-9 A.01 2

S DFX (modified), failure of one fan 1.2x10-8 A.02 2

and one spray train S DF. spray failure 1.6x10-8 A.03 2

S DFX, failure of sprays and fans 2.9xlO-ll A.04 2S DF with convective heat transfer 1.6xlO-8 A.05 2

coefficients increased by a factor of 5 A.06 6% hydrogen ignition limit 4.8x10-6 A.13 Upper Plenum Ignition Failure Standard S D v/35% Zr reaction 4.8x10-6 B.00 2

C.00 CLASIX Combustion Parameters 4.8x10-6 w/TVA source term C.01 HECTR Combustion Parameters 4.Bx10-6 w/C.00 source term C.02 COMPARE Combustion Parameters 4.Br10-6 w/C.00 source term D.00 100% Zr reaction, core melt 4.8xlO-6 D.01 Containment vented D.02 Partial oxygen depletion E.00 37% Zr reaction v/ partial core melt 4.5x10-6 S D Cases (with 75% Zr reacticu unless noted) 3.5xig-6 1

F.01 Partial oxygen depletion 3.5x10-6 G.00 37% Zr reaction 3.5x10-6

Mr. John Larkins January 24, 1984 Table 3 (continued)

Case Description Probability S H Cases (with 75% Zr reaction unless noted) 1.3x10-5 1

Standard S H sequence H.00 1

H.01 Partial oxygen depletion S HF, failure of spray recirculation 3.0x10-6 I.01-1 I.04 1800 see into accident.

Investigates effects of ice condenser modeling parameters S HF, heat transfer coefficients 3.0x10-6 I.05 1

increased by a factor of 5 S HF w/ partial oxygen depletion 3.0x10-6 I.06 1

S HF, 100 t 2r reaction, core melt 3.0x10-6 J.00 1

with vessel breach J.01 J.00 with containment venting J.02 J.00 with partial oxygen depletion S HF, 37% Zr reaction, partial core 3.0x10-6 K.00 1

melt K.01 K.00 with partial oxygen depletion TML cases (with 75% Zr reaction unless noted)

L.01 TMLU, failure of Chemical Volume and 1x10-6 Control System with partial oxygen depletion M.00 TMLB', fan and spray failure, 4x10-7 recovered at 8440 see into accident M.01 M.30 with partial oxygen depletion N.00 TMLB', no recovery, 100% Zr reaction, 1x10-6 core melt, 12% ignition criterion N.01 N.00 with containment venting N.02 N.00 with partial oxygen depletion 0.00 TMLB', 27% Zr reaction, partial core 1x10-6 melt 0.01 0.00 with partial oxygen depletion P.00 TMLB'. 65% Zr reaction, partial core lx10-6 melt P.01 P.00 with containment venting These probabilities were obtained in a private communication 7

with SNLA personnel involved in PRA analyses.

Mr. John Larkins January 24, 1984 S D Cases 2

The S D case with 75 percent zirconium reaction has been 2

widely used as a quasi-standard design basis for hydrogen burn analysis (a small diameter LOCA with failure of emergency core cooling).

This scenario is considered in cases A.00 and C.00.

Both cases indicate acceptable component temperatures (except for the thin-walled component in the upper plenum in C.00).

Comparison of the two cases illustrates the sensitivity to hydrogen source terms.

Case C.00 uses a source term from a TVA analysis while the A.00 case uses a MARCH source term.

The A.00 case has no bu ns in the lower compartment and the C.00 case has one.

Yet, the A.00 case has higher peak surface temperatures.

The somewhat surprising result of high temperatures in the absence of local burns is more dramatic in cases A.01 through A.04 in the lower compartment.

In all four cases, burns in the upper compartment result in temperatures exceeding 440K for the thin-walled component in the lower compartment.

Similar peak temperatures occur for the transmitter cover plate in three of the four cases.

Reference to Table 3 shows that these four cases involve the failure of the air recirculation fans and/or the containment sprays.

This underscores the importance of these engineered safety features (ESFs).

The failure of these ESFs results in a very hot environment in the upper compartment and upper plenum, as evidenced by the high peak temperatures and large temperature increases (some exceeding 200K) there.

The occurrence of high peak temperatures in the absence of local burns is only associated with the few S D cases 2

mentioned.

Cases A.05 and A.03 are similar except that heat transfer coefficients for the A.05 case are increased by a factor of 5: thus, component surface temperatures are higher for this case.

In case A.13 the up; r plenum igniters are assumed to have failed.

Because hydrogen isn't burned in the upper plenum, more is available for combustion elsewhere.

This results in two burns in the lower compartment and higher surface temperatures.

The B.00 case is the same as the A.00 case but with zirconium oxidation limited to 35 percent.

Surprisingly, though less hydrogen is released in the B.00 case, surf ace temperatures in the lower compartment are slightly higher.

This is due to a change in the timing of events, such as ice melting and steam injection, brought about by the way MARCH handles the 35 percent xirconium reaction limit.

Though the temperature differences between the A.00 and B.00 cases are small, the thin-walled component did exceed 440K for the B.00 scenario.

~,'

s Mr. John Larkins January 24, 1984 The C.xx cases use a source term from a CLASIX analysis done by TVA.

Three separate HECTR runs were made using this source term and the combustion parameters from three hydrogen combustion codes: HECTR, CLASIX, and COMPARE.

The caseg and corresponding parameter sets are given in Table 4.

The results a

for these cases indicate moderately differentstemperatures.

Table 4 C.xx Case Combustion Parameters Case Code Parameters C.00 HECTR C.01 CLASIX C.02 COMPARE The D.xx cases are core-melt scenarios.

They assume 100 percent zirconium reaction.

The resulting temperatures in the lower compartment are quite high.

However, it is not possible to separate the effects of hydrogen combustion from the effects brought about by the vessel breach.

The E.xx cases are also core melt scenarios, but the core is assumed to be quenched in some way after 35 percent zirconium reaction and vessel breach.

The only high tempera-ture occurs in the upper plenum for the thin-walled component.

Lower compartment temperatures are noticeably lower than for the D.xx cases.

Several S D cases result in no excessive temperatures.

2 Cases A.06 through A.09 were run varying the hydrogen ignition limit between 6 and 10 volume-percent.

Cases A.10 through A.12 were run to check sensitivity to combustion completeness and flame speed.

Case A.14 investigated the effects of oxygen depletion of the containment atmosphere.

Case A.15 assumes removal cf the ice condenser doors.

As with case A.14 E.02 examines the effects of oxygen depletion.

S D Cases 1

This scenario is the subject of the F.xx and G.xx cases, all of which result in high surface temperatures in the love';

compartment but nowhere else.

The F.xx cases assume 75 percent zirconium reaction.

Though results for the F.00 case are not available, the F.01 case, which investigates the effect of oxygen depletion, indicates very high peak temperatures for the thin-walled model and the transmitter cover plate.

Examinat ion of other cases investigating oxygen depletion shows little difference with the corresponding non-depleted case.

When

• l Mr. John Larkins January 24, 1984 large temperature differences do exist between corresponding surfacec, the higher temperatures normally occur for the,non-depleted case.

Thus, it is reasonable to assume that the temperatures for the F.00 case would be at least as high as a

those of the F.01 case.

The G.00 case assumes 37 percent zirconium reaction.

Though this amount is only half that of the F.xx cases, temperature rises and peak temperatures are comparable.

As with the A.00 and B.00 cases, this is brought about by difterences in the timing of the ice melting and steam injection.

5 H Cases 1

This scenario is covered by the H.xx cases.

Both the base and oxygen depleted cases show high temperatures resulting from large number of burns in the lower compartment.

a S HF Cases 1

This scenario is the S H sequence with the failure of 1

spray recirculation.

It is covered by the I.xx through K.xx The I.xx cases are for a degraded core while the J.xx cases.

and K.xx cases deal with core melt (J.xx for 100 percent zirconium reaction and K.xx for 37 percent).

Comparison of the I.xx and H.xx cases shows generally similar results for lower compartment temperatures.

As expected, the loss of spray recirculation results in generally higher temperatures in the upper compartment and upper plenum.

For the core melt scenarios, temperatures in the lower compartment are about the same.

However, temperatures in the As other two compartments are much higher for the J.xx cases.

expected, due to the smaller amount of zirconium oxidized in there are fewer burns; there are none at all in the K.xx cases, the upper compartment.

THL Cases These scenarios are covered by the L.xx through P.xx cases.

The L.xx and M.xx cases show high temperatures in the upper plenum but not the upper compartment.

Temperature rises in the lower compartment are guite low; however, peat temperatures are high.

This is because the preburn environment in that compartment is such that initial temperatures are near or above 440K.

~. ~

Mr. John Larkins January 24, 1984 The N.xx through P.xx cases deal with core melt (N.xx, 100 percent zirecnium reaction; 0.xx, 27 percent: P.xx, 67 percent).

None of the cases show excessive temperatures in the lower com-There are no burns there.

Temperature rises and partment.

peak temperatures in the upper compartment and upper plenum 4

tenh to follow the amount of zirconium oxidized with the 100 percent reaction scenario resulting in the highest temperatures and the 27 percent reaction the lowest.

There are no burns in the upper compartment for the 0.xx (27 percent reaction) cases.

General Summary One of the more interesting results of the HECTR runs is with few exceptions, the peak temperatures for the thin

that, and thick walled aluminum components bound the peak temperature for the transmitter cover.

Another somewhat surprising result is that, even in the absence of local burns, components can, in some circumstances, reach high temperatures.

For the TMLU and TMLB' cases, surfaces in the lower compartment can exceed 440K prior to burn initiation.

For sequences involving local burns, high peak temperatures

(>440K) are generally the result of temperature rises in excess There of 100K.

However, the converse isn't necessary true.

are several instances in which temperature changes in excess of 100K do not result in high peak temperatures.

It must also be pointed out that there are several instances in which temperature rises of less than 100K lead to high peak surface temperatures.

Significantly, these occur in the lower compartment (see cases C.01, D.00, D.01. F.01, G.00, H.01.

I.06, J.00, J.01, J.02, K.00, K.01, L.01, and M.00).

Some general observations can be made for each compartment for sequences in which burns occur.

The upper compartment is relatively mild for S D scenarios, S D and S H sequences.

2 1

1 It gets hot for the S HF and TMLB' sequences.

In the S HF 1

1 scenario, the sprays have failed and in the TMLB' sequence, loss of power prevents the use of most, if not all, heat removal systems (in the TMLB' sequence, power is eventually restored).

The upper plenum is hot for the S EF, TMLU and I

TMLB' sequences.

The lower compartment, where most of the safety equipment is located, is hot for nearly all sequences except some S D variations and the TMLB' scenarios.

In terms 2

of the number of cases for which the component surface temperatures exceeded 440K, the lower compartment has the most.

i t

I

~..'

Mr. John Larkins January 24, 1984 Finally, the relative probabilities of these events should be addressed.

As mentioned earlier, the S D event with 75 2

percent zirconium rcaction has become the quasi-official standard scenario for the industry when considering hydrogen burns.

The A.00 and C.xx cases analyzing this event indicate gen'erally acceptable surface temperatures.

The approximate probability of such an accident is 4.8x10-6 per reactor year.

However, there are other sequences, with comparable probabilities, which do result in high surface temperatures in the lower compartment.

These are summarized in Table 5.

Table 5 Probabilities of Sequences which Produce High Component Temperatures Secuence Erobability per Reactor Year SD 3.5x10-6 1

1.3x10-5 SH 1

S HF 3.Ox10-6 1

TMLU l.Ox10-6 In view of these probabilities. consideration of theFe sequences may be warranted.

Conclusion The primary conclusion to be drawn from the analyses presented here is that excessive temperatures, i.e.,

temperatures above those at which the operability of equipment has been demonstrated, are the likely result of some accident scenarios which have similar probabilities to the S D 2

sequence which doesn't precipitate such high temperatures.

Thus, we feel that in consideration of equipment survival, it is necessary to demonstrate that either (1) equipment can survive the higher temperature, (2) a spectrum of probable scenarios, at least including those considered here, fail to produce excessive temperatures in the specific plant being treated, or (3) the safety related equipment is adequately Protected from full exposure to the possible hydrogen burns.

We intend to follow this letter with a formal SAND /NUREG report.

Your comments on the content and conclusions of this letter are welcome.

We hope the information here will be helpful in your consideration of the potential consequences of

=.

---

.=..........................

t is a

i 95 )

22 i

si i

t r i ll!

3 13s 3d 1

31 3-33 3 a a l 12 t

e 41 1

d

!!i is 1

11 b

43 i

d G 2 24 9

s i i

!iii d

133 31 1

11 11191 1 1 11 r

3-e la

i did da 3

13 W

e 11 t i i did li ad I

t als?

i id!

31 i

la 3313223 21 d

6 Ed 1

313 dd

's 14 7 I

I ECQ cim e.

-l 2

'v,u Y s o a o al q q 'c q l

t 29di 1

2 33 25

1 1ai s;13 :23 5 r 1913 1

31J 13 1

33 3 s1 1459 13 214113 13!ld; i

1 1 id3issa r 3 s

e e

2d!?

d 1da 29 1 9 1s !1: 12 21

[= 7 3333 1 339 il 4

33 3 di i f2i 13 pi s

i sle d

dai '11 ai i#; ;'i ti al i

t 21da 1

ddi 4d Ei a BJ Ha di 31 i

1211 1

822 '30 2

15 i d541 -91 29

• I s!

?

41!i dd!

'i d a

31 11/ !i a ?

11 l

4 l

."g _JJ2 4w ee 4-q u

.,ju u s J p

o -

. 5 l

t 54

'9 idi 1

11 Edi ;

21 i

a G 4343 3 1sd 29 4

al 113 3511 di W

e 1

9 di s 1

ti s 1 11 91 i

t i 1d

! 1 did 3d da li s a ja i

t t h3423 3 193 55 n1 94 15 13" B1 31 S

e i d1 fd 191 11 14 1 4 44 9s t 3 1 #

3 i Hi s

! d Jd i 138

12 l -e r !!31si l 31f i 5 D i i

s fi ss a?

33

=

I I

e s

1 e

i ld Ni

-5 4

5 733J51111332314J3 i

333 1 33 at I

! l i i

J l

i ls 1

3 1

d is :.,.

y r

y e

?

W

] ; d u e

r i s i

he F $$$3333333333$$3 SJd

$$3 II 3

5 3 1

i a

1 49

li i

s as r

i e

G J

Haa 1 -s 1

a al I

c h2y 13s1Ejf3?Jid31e 3 4

al.a 11 1 ft a

E ai JNr 11 1I I

i t 13 igi i l!

1 d

du

• cldddO'd 0

a qd ca d c.C 0

S a

.s.s e e

c d s df t 192ss123;dAllaaa 15; l e; d2

. 3 3

9; e ri?s1133?flid3E11 a

13d 9 31 di 1

1 Si e Biddis2d13911!11 i ll i ll s i 1

1 li

1. 1995dadad322291 S AR

! ! n 22 2

2 92 r

d r 133131231iid!EG1 8

14d li a is 4

a 1 1 l

e

/

e 14i 11 d1145319: 11 141 da#

EE 1

1 i s c 112s14124dd)35 s a 21 4 11 1 a 1 s

E E i

i+@y 3;951#311193s251i did ill 5d A d1 3 l

D 3321s5 11s!3a3di a 911 i ff di 2

d 1;

e f, f dhM C

d M W e d r e d U

c C $ E a

r e J

U 61 r s

J t t#J114124;1411E1 a

e 1; 3 42 3

a r33d233!!al3dli g; g

31; 311 ad a

2 is m

b 19?d3alill2113il 1

1 111 12 3d 1 1 r

Y h

SE YEE E Y

b b

b k

v a

e

! o r i ga 3 sd ; 19114 d a d i i

s i s i dd 13 12 1 l i J

sli gdas s 11i 12111 i

i 111 la

1 1 1 e

J F J145 1 5 ; ; 2 1 ; 2 dja ;

1 a

E e 1 W3 U

s uyai 31 i giii 1 3s i Ed3 Hi i !

34 42 i si

~.'

Mr. John Larkins January 24, 1984 bydrogen burns.

Please call us for any aaditional information or clarification you need.

Sincerely.

Q

.'. & \\.

~.

Vincent J.

ndini ClIG D) h' A

\\

W. H. McCulloch Safety Systems Assessment Division 6445 VJD:WHM:6445:bjt:2050 Copy to:

USNRC W. Farmer USNRC K. Parczewski USNRC H. Garg USNRC C. Tinkler 6411 J.

Linebarger 6427 M. Berman 6427 A.

Camp 6440 D. Dahlgren 6445 B.

Bader 6445 W. McCulloch 6445 V. Dandini

~

4 1

a]

[ -.

e....

le*a :-

s r

i, o.m e

f ROUTING AND TRANSMITTAL SLIP

!memnumu M

marsmans Tm m.m.. emee siebor room ni.neer.

innt.m cat.

L 6uisses, gency/ Peso H. Garg, EQB/DE g,

i

~

K. Parczewski, CEB/DE 3.

C. Tinkler/R. Palla CSB/DAI s.

N. SU GIB/ DST 4,

P. Worthington, CHEB/DET/RES g

Action Fue Note and Retum Approval For Descance Per Conversation As Requested For Correction Prepare Repy Deculate For Your taformation See Me comment investigste SJgnature Coortfinetton Justly stowtxs Attached is an infomal draft report from SANDI A on Equipment Survival in an Ice Condenser Plant.

Please review and provide marked-up or written -

coments by COB March 28, 1984 DO NOT use this form as a RECORD of approvsts, concurrences, disposats, casarances, and simitar actions TROtt. (Name, arg spetW, Agong/ Posy Room No.-Sleg.

~

126M John T. Larkins

' i Phone me.

t r

x74478 I

N orTewAt ronu 41 (Rev. 7-75) e=w e

i OE MI U III*I 1

• CPf.: 1993 0 - 383-179 f2321 M6

ROUGH DRAFT date:

Tebruary 22. 1984 meno to:

John Larkins. USNRC/RE5 Bill Farmer, U5NRC/RE5 from:

V.

H.

McCu))och, SNLA subject:

Ice Condenser Equipment Surv2vab222ty keport Here 25 a copy of the Sequoyah Report as it is scing 2nto our slan-off process.

A::ording to the current procecures, after it is s2gned-off here, it w211 be formally submitted to you for comments before the f2nal draft is prepared, given a NUREG nunter, and distributed.

Could you review this draft so that we might 2n:orporate your ctaments in the next rewrite?

I think this would signif2cantly speed up the process.

Since it hasn t been fully reviewed, this draft shouldn't be distributed further.

t f,f.h[

'~ %. h.,'J, J g Q If,

r 1. <

DRAFT PRELIMINARY REPORT FOR COMMENT FIN NO. A 1270 a

HYDROGEN BURN SURVIVAL PROGRAM HECTR ANALYSIS OF EQUIPMENT TEMPERATURE RESPONSES TO HYDROGEN BURNS IN AN ICE CONDENSER CONTAINMENT Prepared by Sandia National Laboratories Albuquerque, NM 87185 Operated by Sandia Corporation for the U.S. Department of Energy U.S. Nuclear Regulatory Commission NOTICE THIS DRAFT PRELIMINARY REPORT IS ISSUED ONLY TO PARTICIPANTS IN THE DESIGNATED COOPERATIVE PROGRAM This report was prepared in contemplation of Commission action.

It has not received patent review and may contain information received in confidence.

Therefore, the contents of this report should neither be disclosed to others nor reproduced, wholly or partially, unicss written permission to do so has been obtained from the appropriate USNRC office.

The recipient is requested to take the necessary action to ensure the protection of this report.

c;

seWP Yf-/

//

8 s

HECTR ANALYSIS OF EQUIPMENT TEMPERATURE RESPONSES TO SELECT HYDROGEN BURNS IN AN ICE CONDENSER CONTAINMENT Vincent J. Dandini William H. McCulloch February 1984 Sandia National Laboratories Albuquerque, NM 87185 Operated by Sandia Corporation for the U.S. Department of Energy Prepared for Severe Accident Assessment Branch Division of Accident Evaluation Office of Nuclear Regulatory Research and Chemical Engineering Branch Division of Engineering Office of Nuclear Reactor Regulation Washington, DC 20555 Under Interagency Agreement DOE 40-550-75 NRC Fin Nos. A-1270 and A-1306 G

9

I e

ABSTRACT The HECTR computer code was used to calculate the temperature response of three generic surfaces representative of reactor safety equipment in each of three 3

The locations in an ice condenser containment building.

results of these accident sequences are summarized.

For the S D accident sequence, the calculation does not project 2

However, for other sequences having excessive temperature.

probabilities of occurrence near that of the S D event, 2

component surface temperatures higher than LOCA qualification guidelines are indicated.

M

-iii-

e CONTENTS Page Section 1

Executive Summary.

5

1.0 INTRODUCTION

7 2.0 SURFACE DESCRIPTION AND MODEL 8

3.0 SCENARIOS 10 4.0 RESULTS 19 5.0 GENERAL

SUMMARY

OF RESULTS 34

6.0 CONCLUSION

36 REFERENCES 37 APPENDIX a

-V-

e LIST OF FIGURES P_aSe Figute 5-1 Upper Plenum Transmitter Cover Plate Temperature Response to a TMLB' Event (Case N.02) 21 a

5-2 Upper Plenum Transmitter Cover Plate Temperature Response to an S DPX Event (Case A.04) 22 2

5-3 Upper Plenum Transmitter Cover Plate Temperature Response to an S 3F Event (Case J.00) 23 1

4~

Upper Compartment Transmitter Cover Plate 7-p

• 24 5-4 Response to a TMLB' Event (Case N.02) 5-5 Upper Compartment Transmitter Cover Plate Temperature Response to an 5 HF Event (Case J.00).

26 1

5-6 Upper Compartment Transmitter Cover Plate Temperature Response to an S DFX 2

27 Event (Case A.04) 5-7 Lower Compartment Transmitter Cover Plate Temperature Response to an S D 28 I

Event (Case G.00) 5-8 Lower Compartment Transmitter Cover Plate Temperature Response to an S H 1

29 Event (Case H.01) 5-9 Lower Compartment Transmitter Cover Plate Temperature Response to a TMLU 30 Event (Case L.01) 5-10 Lower Compartment Transmitter Cover Plate Temperature Response to an S D Event with Upper 2

31 Plenum Ignitor Failure (Case A.13)

-vi-

e LI S' OF TABLES Page Ta ble 3-1 Accident Sequence. Event Identification 8

3-2 HECTR-Ice Condenser Analysis Basic 9

Accident Sequences.

4-1 Accident Cases Refulting in High 11 Peak Temperatures 15 4-2 C.xx Case Corbustion Parameters 6-1 Probabilities of Sequences Whien Produce 34 High Component Temperatures 1

N

-vii-

EXECUTIVE

SUMMARY

During certain types of severe nuclear reactor accidents, the possibility exists for the release of large amounts of hydrogen into the reactor containment building.

concern that this hydrogen could build up to concentrations which, if ignited, could cause pressures high enough to threaten containment integrity, has led to the installation, in many reactor buildings, of deliberate ignition systems.

The purpose of these systems is to burn the hydrogen at lower, nonthreatening concentrations.

These deliberate ignitions, while they might prevent the overpressurization of the reactor building, result in thermal environments which coul6 compromise the operation of safety tquipment necessary to monitor the reactor and maintain it in a safe condition.

The HECTR computer code has been developed at Sandia to model the pressure-temperature environments re ulting from the combustion of hydrogen in reactor containment buildings.

Recently, the code was modifisd to include models of three generic surfaces representauive of safety equipnent inside containment to estimate the response of The equipment exposed to the hydrogen burn environment.

surfaces are:

thin aluminum (.125 inch), thick aluminum

(.67 inch), and the.25-inch thick steel cover plate from a Barton pressure transmitter.

For this report, each was modeled at three locations inside an ice condenser

containment:

the ice condenser upper plenum, the upper compartment, and the lower compartment.

Four basic accident sequences with variations on each were considered.

The sequences were:

S D' 8 D, S H, and TML (sequence 2

1 y

nomenclature is that used in the Reactor Safety Study Methodology Applications Program, symbols are summarized in Table 3-h.

Results are given in the Appendix of this report.

The S D sequence with 75 percent zirconium oxidation 2

is widely used as a quasi-standard case for hydrogen burn analyses.

The HECTR analysis of this sequence indicates that component surface temperatures will not reach 440 K (a typical maximum LOCA qualification temperature).

However, variations of this sequence, notably S DX, S DFX, and 2

2 S DF, produce high temperatures in all three compartments.

2 The S D and S H sequences result in high-peak y

y surface temperatures and large temperature rises in the lower compartment but not in tho upper plenum and upper compartment.

Variations of the S HF sequence produce high surface y

~

temperatures in all compartments.

The surface temperatures in the lower compartment for this sequence are very high.

The thin aluminum surface reaches 780 K in some cases and the steel transmitter cover plate exceeds 600 K.

The TMLU and TMLE scenarios result in surface tempera-tures in the lower compartment which exceed 440 K prior to hydrogen ignition.

scenarios produce no excessive temperatures in 4

The TMLB' the lower compartment.

However, these cases do indicate high surface temperatures in the upper plenum and upper compartment.

overall, A review of the analysis results showr that, the lower compartment is more susceptible to high-peak temperatures than the other two compartments which were analyzed.

In large measure, this is due to the hot preburn environment in that location.

Surfaces in the lower compartment can be ex70 sed to high temperature environments for an hour or longer immediately prior to the first hydrogen ignition.

Thus, the surfaces are already at high This high temperatures when the hydrogen starts to burn.

starting temperature and the large temperature rises due to hydrogen combustion result in temperatures which can exceed LOCA qualification guidelines.

Consideration was also given to the approximate probabilities of sequences which produced high temperatures afety in the lower compartment where much of the equipment would be located.

Four sequences (S D, S H, y

y S BF, and TMLU) have probabilities of occurrence which are y

near that for the S D event which produces no high y

Thus, when considering equipment survival in temperatures.

a specific plant, it should be demonstrated that:

(2) a (1) equipment can survive the higher temperature, spectrum of probable scenarios fail to produce excessive temperatures, or (3) equipment is adequately protected from exposure to the hydrogen burn environment.

e e

m 0

2 e

4 e

O M

f.~.. *

1.0 INTRODUCTION

g, One of the many events attendant to the accident at Three Mile Island in March of 1979 was a single hydrogen

(

/

burnI which caused the pressure in the Unit 2 containment building to rise rapidly to about 193 kPa (28 psi).

The integrity of the TMI containment was not compromised, but, the pressure spike at TMI has raised concerer that a similar event in another reactor building might threaten the ability of that structure to isolate fission products from the environment.

One method proposed to prevent such an event is the deliberate ignition of the hydrogen before it accumulates to concentrations which, if burned, might threaten containment integrity.

In this way large pressure pulses can be avoided.

However, the thermal environment resulting from this deliberate ignition may have the undesirab.le side da=3:p 3

~

effect of de5:--tag equipment necessary to maintain the reactor in a safe shutdown condition and monitor plant conditions.

One of the objectives of the Hydrogen Burn Survival (HBS) program at Sandia National Laboratories is to characterire the hydrogen burn environment and its effects on equipment inside the reactor building.

Given the difficulties, expense, and time involved in experimentally characterizing the thermal environment,

, computer codes offer a fast and relatively inexpensive means of describing the conditions and effects resulting from the deliberate ignition of hydrogen in a containment 5uilding.

J

_. -... u.

gECTL The Hydrogen Event:

Containment Transient Response)1EG'!TT' 2

code was developed for this purpose.

Using MARCH to provide the hydrogen and water source terms, HECTR models the pressure-temperature environment in the centainment.

a Recently, the code was modified to include models of three generic surfaces representative of safety equipment at each of three locations inside an ice condenser containment building.

This report describes the HECTR-calculated temperature responses of the representative surfaces when exposed to some accident scenarios which resulted in hydrogen burns.

The pressure environments trom these events are presented in detail, along with a description of HECTR, by Camp et al.#

e e

2.0 SURFACE DESCRIPTION AND MODEL The containment locations considered in the equipment survivability analysis were:

theuppercompartmentg(Ia large open volume where burns involving relatively large e

amounts of hydrogen might occur; the upper plenum of the ice condenser c'where a large number of burns in a small volume

)

might occur; and the lower ccmpartment y!where the steam and -

hydrogen are most likely to be introduced into the containmentbuildingandwheremostofthereacto[ssafety equipment is located.

All surfaces were modeled as one-dimensional slabs, insulated on the back, using a simple transient, implicit, finite difference calculational technique.5 thin aluminum (.125 inch),

The surface types were:

thick aluminum (.67 inch), and the.25-inch thick steel cover plate from a Barton pressure transmitter.

~

..........ma.__--

3.0 SCENARIOS The scenario nomenclature follows that of the Reactor Safety Study Methodology Applications Program (RSSMAP).

Event symbols are identified in Table 3-1.

Table 3-1 6

Accident Sequence Event Identification Event Symbol Intermediate break loss-of-coolant accident S1 (LOCA) (2-6 inch diameter) in reactor coolant system pressure boundary Small break LOCA (2 inch or smaller diameter)

S2 in reactor coolant system pressure boundary T

Transient Failure of emergency core-cooling system D

(ECCS) injection Failure of ECCS recirculation H

Failure of containment spray recirculation F

Failure to maintain coolant invertory in the M

steam generators and transfer heat to the environment using the auxiliary feed water system and secondary steam relief Failure of the chemical volume and control U

system in high-pressure injection mode Loss of power with failure to recover within 1 B'

to 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> of the initiating event s'

Loss of power with recovery Failure to maintain coolant inventory in steam L

generators and transfer heat to the environment using the power conversion system and secondary stea.n relief 8-

Fcur basic scenarios, with variations on each, were considered in the analysis reported here.

These scenarios are summarized in Table 3-2.

Variations to these basic scenarios involved changes in combustion parameters a

tr ::TOTk as well as additional events in the sequences With few exceptions, tue ignition criterion for themselves.

hydrogen in the analysis was set at 8 volume-percent.

Table 3-2 HECTR Ice Condenser Analysis Basic Accident Sequences Description Secuence Small brerA LOCA with ECCS failure SD 2

Intermedicte break LOCA with ECCS failure SD 1

Intermediate break LOCA with f ailure of ECCE S1H recirculatien Transient with failure to maintain coolant TML inventory in steam generators and transfer heat using power conversion system, auxiliary feed water system and secondary steam relief

~.

.a_-._...-..-------..

4.0 RESULTS The results of the surface temperature calculations are given in the Appendix of this report.

Three conditions are presented:

initial temperature (T ), peak temperature g

- T ).

p), and the temperature increase (d T = T (T

p g

The initial temperature is the temperature at the initiation of the first burn in the compartment in which the surface is located.

When no values for T and dT are given, no g

burns occurred in the compartment and T is the result of p

the LOCA environment and burns elsewhere in the containment.

Particular attention was given to those cases which project temperatures of 440 K (332'F) or higher.

This temperature is typical of maximum LOCA qualification temperature guidelines for safety equipment; higher temperatures exceed the level to which operability has been demonstrated.

Those cases which resulted in peak temperatures of 440 K or grea;er are described in Table 4-1.

When obtainable, approximate probabilities per P

rcactorgare also given for those cases.

These probabilities refer only to the accident sequence and are independent of such factors as the amount of zirconium reaction 2nd code parameters.

.S D cases _

2 The S D case with 75 percent zirconium reaction has 2

been widely used as a quasi-standard basis for hydrogen burn analysis (a small diameter LOCA with failure of emergency

. s..=_

.. ~... _..,.

o.......

Table 4-1 Accident Cases Resulting in High-Peak Temperatures (with hydrogen ignition limit set at 8% unless specified)

Case Description Probability 4

S D cases (with 75% Zr reaction unless specified) 4.8 x 10-6 2

S DX, failure of air recirculation fans 8.6 x 10-9 A.01 2

S DFX (modified), failure of one fan 1.2 x 10-8 A.02 2

and one spray train S DF, spray failure 1.6 x 10-8 A.03 2

S DFX, failure of sprays and fans 2.9 x 10-11 A.04 2

S DF with convective heat transfer 1.6 x 10-8 A.05 2

coefficients increased by a factor of 5 A.06 6% hydrogen ignition limit (All other 4.8 x 10-6 cases 8%)

A.13 Upper Plenum Ignitor Failure Standard S D w/35% Zr reaction 4.8 x 10-6 B.00 2

C.00 CLASIX Combustion Parameters 4.8 x 10-6 w/TVA source term C.01 HECTR Combustion Parameters 4.8 x 10-6 w/C.00 source term C.02 COMPARE Combustion Parameters 4.8 x 10-6 w/C.00 source term D.00 100% Zr reaction, core melt 4.8 x 10-6 D.01 Containment vented D.02 Partial oxygen depletion E.00 37% Zr reaction w/ partial core melt 4.8 x 10-6 SID Cases (with 75% Zr reaction unless noted) 3.5 x 10-6 F.01 Partial oxygen depletion 3.5 x 10-6 G.00 37% Zr reaction 3.5 x 10-6 S H Cases (with 75% Zr reaction unless noted) 1.3 x 10-5 _

3

.c Standard S H sequence 1.0 K lo,

H.00 1

H.01 Partial oxygen depletion g,e w so *

~~

S HF, Jailure of spray recirculation 3.0 x 10-6 I.01-1 I.04 1800 see into accidert.

Investigates e

effects of ice condenser modeling parameters S HF, heat transfer coefficients 3.0 x 10-6 I.05 1

increased by a factor of 5

^

............a.a~

Table 4-1 (Continued)

Description Probability Case 3.0 x 10-6 S HF w/ partial oxygen depletion I.06 1

S HF, 100% Zr reaction, core melt 3.0 x 10-6 J.00 1

with vessel breach J.01 J.00 with containment venting J.02 J.00 with partial oxygen depletion S HF, 37% Zr reaction, partial core melt 3.0 x 10-6 K.00 1

K.01 K.00 with partial oxygen depletion TML cases (with 75% Zr reaction unless noted)

L.01 TMLU, failure of Chemical Volume and 1 x 10-6 Control System with partial oxygen depletion 4 x 10-7 M.00 TMLB', fan and spray failure, recovered at 8440 see into accident M.01 M.00 with partial oxygen depletion N.00 TMLB', no recovery, 100% Zr reaction, 1 x 10-6 core melt, 12% ignition criterion N.01 N.00 with containment venting N.02 N.00 with partial oxygen depletion 0.00 TMLB', 27% Zr reaction, partial core melt 1 x 10-6 0.01 0.00 with partial oxygen depletion P.00 TMLB', 65% Zr reaction, partial core 1 x 10-6 melt P.01 P.00 with containment venting y,-

. ~.. ' - - - - - - - - -

.w...............

core cooling).

This scenario is considered in cases A.00 and C.00.

Both cases indicate acceptable component temperatures (except for the thin-walled component in the upper plenum in C.00).

Comparison of the two cases illustrates the sensitivity to hydrogen source terms.

Case 0

C.00 uses a source term from a TVA analysis while the A.00 case uses a MARCH source term.

The A.00 case has no burns in the lower compartment and the C.00 case has one.

Yet, the A.00 case has higher peak surface temperatures.

The somewhat unexpected result of high temperatures in the absence of local burns is more dramatic in cases A.01 through A.04 in the lower compartment.

In all four cases, burns in the upper compartment result in temperatures exceeding 440 K for the thin-walled component in the lower compartment.

Similar peak temperatures occur for the transmitter cover plate in three of the four cases.

Reference to Table 4-1 shows that these four cases involve the failure of the air recirculation fans and/or the containment sprays.

This underscores the importance of these engineered safety features (ESFs).

The failure of these ESFs results in a very hot environment in the upper compartment and upper plenum, as evidenced by the high-peak temperatures and large temperature increases (some exceeding 200 K) in that location.

The occurrence of high-peak temperatures in the absence of local burns is only associated with the few S D cases mentioned.

Cases A.05 2

and A.03 are similar except that heat transfer coefficients

~.'

m.-..__.

for the A.05 case are increased by a factor of Sr thus, component surface temperatures are higher for this case.

In case A.13 the upper plenum igniters are assumed to have failed.

Because hydrogen is not burned in the upper plenum, more is available for combustion elsewhere.

This results in two burns in the lower compartment and higher surface temperatures.

The B.00 case is the same as the A.00 case but with zirconium oxidation limited to 35 percent.

Surprisingly, though less hydrogen is released in the B.00 case, surface temperatures in the lower compartment are slightly higher.

This is due to a change in the timing of events, such as ice melting and steam injection, brought about by the way MARCH handles the 35 percent zirconium reaction limit.

Though the temperature differences between the A.00 and B.00 cases are small, the thin-walled component did exceed 440 K for the B.00 scenario.

The C.xx cases use a source term from a CLASIX ana) '31s done by TVA.

Three separate HECTR runs were made using this source term and the combustion parameters (ignition limits, flame speed, and combustion completeness) from three

~

and COMPARE.'

hydrogen combustion codes:

HECTR, CLASIX, The cases and corresponding parameter sets are given in Table 4-2.

The results indicate moderately different temperatures.

~.'

Table 4-2 C.xx Case Combustion Parameters Case Code Parameters C.00 HECTR C.01 CLASIX C.02 COMPARE The D.xx cases are core-melt scenarios.

They assume 100 percent 2irconium reaction.

The resulting temperatures in the lower compartment are quite high.

However, it is not possible to separate the effects of hydrogen combustion from the effects brought about by the vessel breach.

The E.xx cases are also core-melt scenarios, but the core is assumed to be quenched after 35 percent zirconium reaction and vessel breach.

The only high temperature occurs in the upper plenum for the thin-walled component.

Lower compartment temperatur4n are noticeably lower than for the D.xx cases.

Several S D cases result in no excessive f

2 p

~(

Cases A.06 through A.09 were run varying the temperatures.

hydrogen ignition limit between 6 and 10 volume-percent.

Cases A.10 through A.12 were run to check sensitivity to combustion completeness and flame speed.

Case A.14

. investigated the effects of oxygen depletion of the containment atmosphere.

Case A.15 assumes removal of the ice condenser doors.

As with case A.14, E.02 examines the effects of oxygen depletion.

..........w-..:

5 D Caoes 3

This scenario is the subject of the F.xx and G.xx cases, all of which result in high surface temperatures in the lower compartment but nowhere else.

The F.xx cases assume 75 percent zirconium reaction.

Though results for the F.00 case are not available, the F.01 case, which investigates the effect of oxygen depletion, indicates very high-peak temperatures for the thin-walled model and the transmitter cover plate.

Examination of other cases investigating oxygen depletion shows little difference with the corresponding nondepleted case.

When large temperature differences do exist between corresponding surfaces, the higher temperatures normally occur for the nondepleted Thus, it is reasonable to assume that the case.

temperatures for the F.00 case would be at least as high as those of the F.01 case.

The G.00 case assumes 37 percent zirconium reaction.

Though this amount is only half that of the F.xx cases, temperature rises and peak temperatures are comparable.

As with the A.00 and B.00 cases, this is brought about by differences in the timing of the ice melting and steam injection.

8 H Cases 1

Both the This scenario in covered by the H.xx cases.

base and oxygen depleted cases show high temperatures resulting from a large number of burns in the lower compartment.

un......-......-s-....

S HF Cases y

This scenario is the S H sequence with the failure of y

spray recirculation.

It is covered by the I.xx through K.xx The I.xx cases are for a degraded core while the cases.

J.xx and K.xx cases deal with core melt (J.xx for 100 percent zirconium reaction and K.xx for 37 percent).

Comparison of the I.xx and H.xx cases shows generally similar results for lower compartment temperatures.

As expected, the loss of spray recirculation results in generally higher temperatures in the upper compartment and upper plenum.

For the core-melt scenarios, temperatures in the lower compartment are about the same.

However, temperatures in the other two compartments are much higher for the J.xx As expected, due to the smaller amount of zirconium cases.

oxidized in the K.xx cases, there are fewer burns; there are et all in the upper compartment.

naaa TML Cases These scenarios are covered by the L.xx through P.xx put_Q (rut %

'*S*S-The L.xxgand M.xx cases show high temperatures in the upper plenum but not the upper compartment.

Temperature rises in the lower compartment are quite low; however, peak temperatures are high.

This is because the preburn environment in that compartment is such that initial temperatures are near or above 440 K.

8 T'

...m (MTMLE The N.xx through P.xx cases deal with core melt (N.xx, a

100 percent zirconium reaction: 0.xx, 27 percent: P.xx, 67 percent).

None of the cases show excessive temperatures in the lower compartment.

There are no burns there.

4 Temperature rises and peak temperatures in the upper compartment and upper plenum tend to follow the amount of zirconium oxidized with the 100 percent reaction scenario resulting in the highest temperatures and the 27 percent reaction the lowest.

There are no burns in the upper compartment for the O.xx (27 percent reaction) cases.

0 '

.~

5.0 GENERAL

SUMMARY

OF RESULTS One of the more interesting results of the EECTR runs is that, with few exceptions, the peak temperatures for the thin and thick-walled aluminum components bound the peak A somewhat temperature for the transmitter cover.

unexpected result is that, even in the absence of local burns, components can, in some circumstances, reach high For the TKLU and TML5' cases, surfaces in the temperatures.

lower compartment can exceed 440 K prior to burn initiation.

For sequences involving local burns, high-peak temperatures are generally the result of temperature rises in excess of 100 K.

Sc, even if equipment temperatures have not risen to LOCA qualification levels as a result of the LOCA, hydrogen burns can be sufficiently energetic to push The converse peak temperatures well over that temperature.

is not necessarily true.

There are several instances, particularly in the upper plenum, in which temperature rises of 100 K do not culminate in high-peak temperatures (gases entering the upper plenum are cooled by the ice condenser prior to entry).

It is also significant that instances of temperature rises of less than 50 K can result in high surface temperatures.

This occurs, with one exception, exclusively in the lower compartment where components are j

more likely to be heated by incoming steam to maximum temperatures associated with a LOCA.

1 T.-

Some general results can be stated for each compartment et - - _ n.

for sequences in whic h bens occur.

Consider M thm [

}AJ

~

i

@ face temperatures prior to the first burn [(

Kg,, a Figuf es 5-1

\\

g through 5-3 show the temperature response of the transmitter cover plate as modeled in the upper plenum for three

(_ accident scenarios which result in high-peak temperatures.

At the start of each of the sequences (time = 0) the temperature of the plate is near that of the ice in the ice condenser below the plenum.

Of all the scenarios which result in a peak temperature greater than 440 K for the cover plate in the upper plenum, case N.02, a TMLB' event, shown in Figure 5-1 has the lowest value of T.

From the g

start of the accident the temperature rises slowly from 280 K to 295 K (15 K in about 9300 sec).

Figure 5-2 shows the temperature profile of the plate for an S DFX event, 2

Case A.04.

The value of T for this case is intermediate g

and so is the rate of temperature rise prior to the first burn (22 K in about 4200 sec).

Case J.00 has the highest initial temperature for those scenarios with high-peak cover plate temperatures.

Figure 5-3 shows that the S BF event y

results in a relatively rapid temperature rise prior to the burn (58 K in 3300 sec).

Corresponding results are shown for the transmitter cover plate in the upper compartment.

Case N.02 has the lowest initial temperature for cases having a cover plate peak temperature greater than 440 K.

Figure 5-4 shows a slow rise followed by a slight decline over 3000 seconds to i

i i

i i

i i

i i

i i

e i_

460 _,

440 420

.c

.t e

x 400

.en 380 o

t.u n: 360 f

cc 340 320 300

~

1 % gi y

I I

I

'~

280 0

2000 4000 6000 8000 10000 12000 TIME (seconds)

Figure 5-1.

Upper Plenum Transmitter Cover Plate Temperature Response to a TMLB' Event (Case N.02)

.:..: a 2 -

3 550 g

500 1

q x450 e

e e

c,m 3400 W

e:s 4

m350 w

G.

3 w

300 N

Tg

~

I I

I I

I 250 0

2000 4000 6000 8000 10000 12000 TIME (seconds)

Figure 5-2.

Upper Plenum Transmitter Cover Plate Temperature Respotse to an S DFX Event (Case A.04) 2 500

_i i

i i

i i

i i

i iigiii;iii_

475 T

F 450 T

E 425 g

e e400 e

u a

e f

t 375 T

m cc m 350 N

b f

5 a;

m.T W 325 i

a lE w

F 300

~

275

~I I

I

' ~

250 0

1000 2000 3000 4000 5000 6000 TIME (seconds)

I Figure 5-3.

Upper Plenum Transmitter Cover Plate Temperature Response to an S HF Evcnt (Case J.00) 1

525

_i i

i g

i i

i i

iig i

i i

g i

i g

i i

i_

500 475

.e_

g

~

~

• 450 u

o425 e

o

~

E400 3

g<

@ 375 n.

E w

F 350 325

,-T f

g I

I I

I

~

300 O

2000 4000 6000 8000 10000 12000 TIME (seconds)

Figure 5-4.

Upper Compartment Transmitter Cover Plate Temperature Response to a TMLB' Event (Case N.02)

As was the case for the upper plenum, Case J.00, an T.g and an intermediate S BF sequence has an intermediate Tg y

The cover rate of preburn temperature rise (Figure 5-5).

plate temperature response for the S DFX event is shown in 2

This case again exhibits a relatively rapid Figure 5-6.

rise to T.

g A substantial difference exists between the preburn temperature profiles for the cover plate in the upper plenum and upper ecmpartment and the profiles for the same plate in This difference is significant the lower compartment.

because most of safety equipment necessary to maintain the reactor in a safe configuration and monitor its condition is located in the lower compartment.

Figure 5-7 shows the temperature profile of the cover plate for an S D event (Case G.00) limited to 37 percent y

zirconium reaction.

The plate temperature rises very rapidly to about 390 K at the start of the accident and drops only slightly thereafter.

At the start of the first burn it is rising again.

Figure 5-8 shows a similar profile for an S H event (Case H.01).

~

y The profile for a TMLU event (Case L.01) is shown in The temperature rise to the plateau is slower Figure 5-9.

but there is a significant increase prior to the first burn.

In terms of preburn temperature behavior, Figure 5-10 is another striking example of the lower compartment preburn It shows the cover plate temperature environment.

~

500 _i i

i g

i i

i g

i i

i i

i i

i i

i t

i 480 460 i

c e440 W

e

~

$420 e

3400 w

g Z

$380 m

w n.360

'E w

Z 340 g

I 5

320 I

I I

I I

'~

~

300 O

1000 2000 3000 4000 5000 6000 TIME (seconds)

Figure 5-5.

Upper Compartment Transmitter Cover Plate Temperature Response to an S HF Event 1

(Case J.00).

480 i

i i_

i i

i i

i g

i i

i i

i i

i i

i g

i 460

?440 e

M

,420 ee

,400 c,n

.o 5

~

E380

s F

@360 c

3 w

F340 T

I 320 I

I

~

300 0

2000 4000 6000 8000 10000 12000 TIME (seconds)

Figure 5-6.

Upper Compartment Transmitter Cover Plate Temperature Response to an S DFX Ev'ent 2

(Case A.04).

525 i

i i_

I 500 E>475 e

t e450 g

u en e 426 t

j g

~

5400 g

4 cc E 375 T

-~

N 2

1 w

p-350 -

325 -.-

5 I

I I

l

'~

300 O

1000 2000 3000 4000 5000 TIME (seconds)

Figui.e ' 5-7.

Lower Compertment Transmitter Cover' Plate Temperatur.e.5.espo use to,an S D Event I

(Case G.00) 13..

......~--_..............2.

550 iiiiiii iiii i eiiii ieie i ie i ii i

_500 c

2 o

g e

e e450 o

m

,e W

0:

3 f

F400

/

g E

1. N 2

T W

i g

350

^

''I''''''II''

300 O

1000 2000 3000 4000 5000 6000 7000 8000 TIME (seconds)

Figure 5-8.

Lower Compartment Transmitter Cover Plate Temperature Response to an 5 H Event 1

(Case H.01) 4 e

iiiii 550 i,,i3

,ii,,igii iiiiiii 500 q

e g

g

\\

Tg

• 450 en e

t w

w cc

D

$ 400 CC w

aE m

M 350

'IIII

300 0

3000 6000 9000 12000 15000 TIME (seconds)

Figure 5-9.

Lower Compartment Transmitter Cover Plate Temperature Response to a TMLU Event (Case L.01) i i

i_

i i

i i

i i

i i

i i

i i

i i

i i

i i

i l

460 440 -

~

c

~

2 420 -

ex

\\

T T

e 400 I

E e

t T

w 380 w

~

cc

)e

< 360 m

w a,

2wF340 T

m 320 300

~

O 2000 4000 6000 8000 10000 12000 TIME (seconds) i Lower Compartment Transmitter Cover Plate Figure 5-10.

Temperature Response to an S D Event 2

with Upper Plenum Ignitor Failure (Case A.13) '

3 response for an S D event in which the upper plenum 2

ignitors fail (Case A.13).

After quickly reaching an elevated level, the temperature declines somewhat then increases quickly to 406 K just prior to the first burn.

Considering the environments in each of the compartments for all of the accident scenarios investigated, some general observations can be made for each compartment for sequences in which burns occur.

The upper compartment is relatively mild for the S D, S H, and variations of the S D 2

y y

It is hottest for the S HF and TMLB' y

sequences.

In the S BF sequence, the sprays have failed sequences.

y and in the TMLB' sequence, loss of power prevents the use of all except passive heat removal systems (in the TMLB'

case, power is eventually restored).

The upper plenum is hot for the S HF, TMLU, and TMLB' Pequences and variations of the y

The lower S D sequence involving spray and fan failure.

2 compartment, where a large portion of the safety equipment is located, is hot for nearly all sequences except some S D variations and the TMLB' events.

In terms of the 2

number of cases for which the component surface temperatures exceeded 440 K, the lower compartment has the most.

for cases in Temperature rises in the lower compartment, which burns occur, are generally the same as those in the The other two compartments for corresponding surfaces.

principle reason for the predominance of high temperatures (i.e.,

in the lower compartment is the preburn environment the release of superheated steam into the lower compartment during a LOCA) which raises surf ace temperatures of exposed surf aces to high levels prior to hydrogen ignition.

Figures 5-1 through 5-9 also indicate that temperatures can remain high for some time.

They show the transmitter cover plate to remain abcVe 440 K for 500 to 1000 seconds.

In the L.01 case (TMLU, Figure 5-9), the temperature exceeded 440 K prior to the first burn and remained above that level for about 4200 seconds (70 minutes).

An equipment easing whose surface temperature remains elevated for such a long period of time will act like an oven, causing interior component temperatures to reach that of the 7

surf ace.

e

--- ~ ~ - --

6.0 CONCLUSION

As mentioned earlier, the S D event with 75 percent 2

zirconium reaction has become the quasi-official standard scenario for the industry when considering hydrogen bu.rns.

__o._.

y-The A.00 and C.xx cases en:1,;1s@ this event indicate g( terally acceptable surface temperatures.

The approximate

-6 probability of such an accident is 4.8 x 10 per reactor However, there are other sequences, with comparable year.

Mc+ gh surf ace tep.peraturescky#

in hi probabilities, which do result

,., d - k 4 {

f in the lower compartment.

These are summarized in s

Table 6-1.

Table 6-1 Probabilities of Sequences which Produce High Component Temperatures Probability per Reactor Year Sequence 3.5 x 10-6 SD 1

1.0 x 10-5 SH 1

3.0 x 10-6 S HF 1

1.0 x 10-6 THLU In view of these probabilities, the primary conclusion to be drawn from the analyses presented here is that excessive temperatures, i.e., temperatures above those at which the operability of equipment has been demonstrated, are the likely result of some accident scenarios which have similar probabilities to the S D sequence which does not 2

precipitate such high terperat2res.

Consequently, when considering equipment survival, it should be demonstrated

^

that (1) equipment can survive the higher temperature, (2) a spectrum of probable scenarios, at least ine'.uding those considered here, fail to produce excessive temperatures in the specific plant being treated, or (3) the safety-related equipment is adequately protected from full exposure to the possible hydrogen burns.

REFERENCES 1.

H. Alvares, D. Bearson, and G. Eidem, Investigation of Hydrogen Burn Damage in the Three Mile Island Unit 2 Reactor Building, GEND-INF-023, Vol. 1, U.S. Department of Energy, June 1962.

r 2.

A. L. Camp, M. J. Westet, S. E. Dingman, HECTR:

A 4

Computer Program for Modelling the Response to Hydrogen Burns in Corstainments, SAND 82-1964C, presented at the Second International Workshop on the Impact of Hydrogen on dater Reactor Safety, Llbuquerque, New Mexico, October 1982.

3.

R. O. Wooten and H.

I. Avci, MARCH (Meltdown Accident Response Characteristfes) Code Description ano Users Manual, NUREG/CR-1711,~U.S. Nuclear Regulatory commission, Washington, D.C., 1980.

4.

A.

L. Camp, V.

L. Behr, F. E. Haskin, MARCH-HECTR Analysis of Selected Accidents in an Ice Condenser Containment in preparation.

5.

B. V. Karlekkr,R.M. Desmond, Engineering Heat Transfer, West 1977.

6.

D.

D. Carlson et al.,

Reactor Safety Study Methodology Applications Program:

Sequoyah #1 PWR Power Plant, SANDBO-1897, NUREG/CR 1659, bandia National Laboratories and Battelle Columbus Laboratories, February 1981.

fMwrIA Inft:rel wvu.stmusi+n with S. W. Hatch Systems safety Analysis Division 6412.,SNLA Reactor 7.

8.

Letter from L. M. Mills, Tennessee valley Authority to E. Adensam, Director of Nuclear Reactor Regulation, USNRC, December 1, 19 81.

9.

R. G. Gido, A. Koestel, Hydrogen Burn Analysis of Ice Condenser Containments, NUREG/CR-3278, LA-9749-MS, Los Alamos National Laboratory, Draft 1983.

1_.

s...

. ~., -.. _

.- u.., *..

9 APPENDIX RESULTS OF EQUIPMENT TEMPERATURE RESPONSE CALCULATIONS

( kh

=-. _. m L k

6

- 4.i u

i x

di: izi E 5 si Ei i i isz

d. i e '. I.a E.i.

d ea t'i..n d s. d i. e. s. s. e. i s e. i.. isI.s.:i. "m a a ii 4 -

s E

a s.

d 5

I ZiZ ili i E si is i E sii e..

e e

1 c

m a

s:s sas s:

,f:

jj P4555 Ediiiiii~li i sif sii si 8 s si siid E lii i

a-I.n aaa EEs mE g I a.

a a

man a ama n

8 a

a

-e-2-

m

-, em, e --

e 4

8 8

S:8 111 E E 11 El E E 151 l

neig*g...i.ss==.e.=e : gg...,,

g ge giri i. e.=<.a

=

-ed.

s s..

- - - -.:a

...; a a

- E IE.i-i a a

e g

s.E

a i

s a maa sn x

1 stgg;- - ---- ------ - n. -.- -- sm - zu j:=-s:---.

i i di *iti" i isi d-"a stid::is:ii nii six ci xi j

a c

W 1....

Ea,....

i. i:a e::::

ia-2.a a eg s.:g5gsac e,s..g a m.

asa si L.

m-s ses-an ::

3 i i si ise* s sia v,isiis isitisinai s

f'a eiem a ma -

- at - ---

L-n

--.n-e-eneemen-we ~~ ee e e es

---e m --e tw ss:::::::::::=

= ::t :n : : s=

es g e n::

=

5j Pfiiiiiiiiiiliiii E iiE iii Ei i E di ggin g gig E

- EE : ***

}'

in rg-isisi"mseide ig igi igg ti i i is e

s a

---m--

t-3 gs-siggszi::issi

=is ins :- t: if is:s i g-i s

t E

-ee.

-e..mee

.e e--

..e-p,.}I.~ae-igag.a. a die i.gg ::

f. s

- af. i ggd

~

ed 2

ss 6

.,a a RS 23.

s:sLL-a ass R 38..

i n--

---en n -.

n-ew e

en mm.m e a-a

--a

5I c=

a a n-

== : t =a-e =:ces:

-:- gmE II m-a a a -----

t.

a ll :-- :: : n--. --.. - e s e. --...- -... -. -.....

.t-smtsiisiiiiiinidi i i its si a ci is m' it_a a

• -(e e 2a. s 3=.3

--m-ne-ne-e --e ese e.e c

e

- - -- - -- ggg gg = - ez c zn=

- a --

emsx:.as==gn=cta=

z E

en -, --

--.e-ne e m.e.,

ns me e e

51EEEE51 Rest 1RIII I I III II E EE 8E 5 EEE 1

unidiisaisititisi i :

sin ni i di is i gii

r,' 4 f,,,

gii i di i

e

-,, ---- -me is e.ss.

E a.. sse

=ma T

m CfN i is l

uti i si 19

}l Mit i if isti Eii is a

CEE i si usi i si 3 -- - ee een --

j M55 I RE E!if EEE RE i i e

-.i c- -

st I).t at-....-...

i is e ii nii is

-z ss-1 c

m-

- --. e.i,.se m. --

Mi s i:

as S--

a a

na

- -- get sia et in e

• e t a

-z.

m e -- e-e-e me sn c rs 5:g: := 23

!:= Mis i $5 siis sii fi e-e e. -. e.-+ -e axe E== .:c;. Ist m-l c 1 i id Wiid %$i - a -c-a $5*s gea e 2 i

• ** * -* *.- *e

.. s := K:.:. 5=. 6 C .3..

• m *. *
• r==

= se RER-ese AR-A

==- Trp =a.,... at 4 i si igit si I ee e- -e m --- e g ess E EE Essi Esa st l m a a se gEEE se as R RR Am 1 e-e e u

ac =tas sa}}