ML20081L059

Rev 1 to Oyster Creek Plant-Specific Oxygen Generation Following Loca

ML20081L059
Person / Time
Site:	Oyster Creek
Issue date:	06/18/1991
From:	Bond G, Nicholas Trikouros GENERAL PUBLIC UTILITIES CORP.
To:
Shared Package
ML20081L054	List: ML20081L049 ML20081L059
References
TR-081, TR-081-R01, TR-81, TR-81-R1, NUDOCS 9107020338
Download: ML20081L059 (43)

Text

.

c.

OYSTER CREEK NUCLEAR GENERATING STATION TOPICAL REPORT 081, REV.1 OYSTER CREEK PLANT SPECIFIC OXYGEN GENERATION FOLLOWING A LOCA June 18,1991 PREPARED BY:

J. D. DOUGHER R. V. FURIA L C. PO G. R. TAYLOR APPROVED: ]l Y j,}m 1,z,j //zw

~

/

N. G. TRIKOUROS s

G.R. BOND hffo e

38 910624 ck 0300o219 PDR

, TR<>81

-s-Rev. O Page 2 SECTIQN TOPtQ f_AQj A

EXECUTIVE

SUMMARY

4 UST OF FIGURES 6 UST OF TABLES 7

1.0 INTRODUCTION

8 2.0 OBJECTIVES 9 3.0 GENERAL DISCUSSION OF POST ACCID:NT RADIOLYSIS 10 3.1 Temperature and Turbulence Effects for Non-Boiling Water 10 32 Boiling 12 3.3 Impurities 12 40 OYSTER CREEK SPECIFIC IODINE RELEASE AND METAL WATER REACTION 16 4.1 lodine Concentration from LOCA with and Degraded Conditions 16 4.2 Relationship between Metal Water Reaction and lodine Release 18 5.0 PLANT SPECIFIC OXYGEN CONCENTRATION WITH REVISED G VALUES 23 5.1 Methodology 23 5.2 Results 24 5.2.1 Oyster Creek Specife Oxygen Generation 24 5.2.2 Oxygen Concentration Following Severe Accidents 24 5.2.2.1 Total Core 24 5.2.2.2 Localized Effects 25 5.2.2.3 lodine Release Without MWR 26 5.2.3 Additional Conservatisms 27 LCP. GEN

. -m_ _

^

TR 081 Rev.O Page 3 I i

JfSLLOF CONTENTS (cont'd)

SECTION TOPIC M

60 CONCLUSIONS 36

7.0 REFERENCES

37 APPENDIX A: OXYGEN VS TIME CALCULATION METHODOLOGY 38 APPENDIX B: NRC STAFF SAFETY EVALUATION ON NEDO-22155 43 TOTAL PAGES 47 i

l LCF. GEN L

1-

-- - _ _ _ _ . _ _ ___ _ __ _.. _ _ ______ s _ . _ _. ._ _ _ _ _. _ _ _ . ~ .

TR 08t Rev. O Pags ,;

EXECtfTNE

SUMMARY

In Enclosure 2 of their *Clanfication of NRC Staff Position on Hydrogen Mitigation Requirements 10CFR50 44 Oyster Creek Nuclear Generating Station' dated November 6,1990, the NRC staff questioned the radiolytic oxygen I

generation rates used in NEDO 22155 The Staff stateo that the results in NEDO-22tSS were applicable to pure water or water containing only minimal amounts of impurities and that including the effect of iodine could drastically change the results. The Staff also indicated that post accident hydrogen and iodine concentrations may vary during an accident and are specific for each indNidual plant.

In order to respond to the NRC Staff's concems, GPUN has prepared Topical Report 081,

• Oyster Creek Plant Specific Oxygen Generation Fdiowing a LOCA*. This report calculates the oxygen concentration in the OC containment as a function of time following a LOCA and conservatNely accounts for the hydrogen and iodine l

concentrations in the containment water. The methodology described in Appendix A of NUREG 0800 (USNRC SRP Section 6 2 5), ' Combustible Gas Contrd in Containment *, is utilized except that the non boiling oxygen generation rate is calculated as a fun: tion of dissolved iodbe and hydrogen.-

An Oyster Creek plant specific iodine concentration was calculated for both the base case LOCA and for a more severe LOCA event in which core cooling is degraded such that a metal water reaction 5 times that of the base case LOCA occurs. The latter case results in iodine releases that are 300 times more than the base case LOCA. A plant specific fuel heatup calculation, with and without' degraded ECCS performance, was performed to determine fuel rod temperatures, metal-water reactk>n rates and the number of failed fuel rods. The iodine releases were determined by comparing the calculated fuel centertine temperature for the failed fuel rods against NUREG/CR-2367, " Updated Best Estimate LOCA Radiation Signature".

l The resufts of the evaluation show that for the iodine and hydrogen concentrations that would be expected as a asult of a relatively severe event, such as a LOCA wth degraded ECCS performance, the oxygen concentration insido containment would remain below the 5% oxygen flammability limit. For very severe events in which 30% of the LCP. GEN

i TR G1 Rsv. O Pag) 5 core iodine is released and 40% of the core undergoas metal water reaction, the flammable limit is not reached for about a year. Postulated events in which significant amounts of iodine are produced without substantial metal water reaction are not credible.

l I

i l

LCP. GEN

TRM1 l -- Rev. O i

l Page 6

{

US'T OF FIGURES AGUAE If.L5 PAQ) 31 CRNL EXPERIMENT NO 11 14 32 ORNL EXPERIMENT NO.10 15 41 OYSTER CREEK POWER DISTRIBUTION HISTOGRAM 19 42 FAILEO PIN MWR ANO IODINE RELEASE AS A FUNCTION OF TEMPERATURE 22 51 OXYGEN CONCENTRATION VS. TIME FOR BASE CASE LOCA 28 52 OXYGEN CONCENTRATION VS TIME FOR DEGRADED LOCA 29 53 OXYGEN CONCENTRATION VS TIME FOR IODINE = 1.4% : MWR = 2 24% 30 5-4 OXYGEN CONCENTRATION VS. TIME FOR IODINE = 10% : MWR = iS% 31 55 OXYGEN CONCENTRATION VS TIME FOR IODINE = 20% . MWR = 30% 32 54 OXYGEN CONCENTRATION VS. TIME FOR IODINE = 30% : MWR = 40% 33 57 OXYGEN CONCENTR ATION VS. TIME FOR IODINE = 4 26% : MWR = 5.36% 34 LCP. GEN

TR481 Rev.O Page 7 UST OF TABWS TABLE I!ILE PAjj 41 OYSTER CREEK IODINE RELEASE DURING BASE CASE LOCA 20 42 OYSTER CREEK IODINE RELEASE DURING LOCA WITH DEGRADED 21 CORE COOUNG 51

SUMMARY

OF RESULTS 35 i

l I

l LCP GEN

TR481 g

Rev. O Pag) 9 1.0 LNT.EQDUCTION On November 6,1990, the NRC issued a letter to GPUN entitled. 'Clardication of NRC Staff Position on Hydrogen Mitigation Requirementa 10CFR50 44 - Oyster Creek Nuclear Generating Station *, (Ref.1-1). Tne letter had two enclosures: Enclosure 1 stated the Staffs position on BWR Mark I compliance W.th the regulations in geeral, and Enclosure 2 was a Safety Eval' ' tion on the BWR Owner's Group riethodology for determining the oxygen generation rates by radiolytic decomposition (NEDO 22155 Ref.12). The Safety Evaluation disagreed with the NEDO report on the radiolytic gas generation rate for boiling and non-boiling conditions. The data which the NRC Staff used was based on an experiment conducted by ORNL (Ref. 31) for pure water and a theoretical model developed by BNL for water contaminated wrth iodine (Ref. 3 2).

Both have shown gas production rates higher than the NEDO assumed values. In this report, GPUN will use the NRC recommended model, with consideration of beyond design basis post accident condrtions for both boiling and non-boiling reactor coolant water to calculate the Oyster Creek plant specific oxygen concentration. In particular, the iodine release fraction for conditions complying with the 10CFR50 44 requirements for degraded ECCS performance and its impact on the oxygen production rate will be addressed.

LCP GEN

s VR 081 Rev. O PQge 9

20 OBJECTIVES The pnmary objectives of this report are as follows:

a) To utilize the basic methodology provided by the NRC in NUREG 0800 (SRP Section 6 2 5),

Appendix A (Ref. 51) for calculating combustible gas concentrations in containment, with modtfications to account for the effect of dissolved lodir'e and hydrogen on the radiolytic gene'ation rate (G value),

b) To develop a value for the concentration of iodine in the containment water following a large break LOCA with a degradation of the ECC system such that the resulting metal water reaction (and resulting hydrogen release) is 5 times that resulting from a base case LOCA (without ECCS degradation).

c) To deterrnine the Oyster Creek plant specific containment oxygen concentration as a function of time for the degraded ECC system performance condition evaluated and for more severe accidents as well.

d) To show that inerting is effective in preventing a flammability condition in the containment following a relatively severe accident in which iodine is released from the core and significant metal-water reaction occurs.

l l

l l

l LCP. GEN l

+

TR481 I' Rev.O Pabe 10 3.0 . ' GENERAL DISCUSSION OF POST ACCIDENT RADICLYSIS For post accident radiolytic decomposition of water, Regulatory Guide 1.7 recommends that G(09 )=0.25 be used for both boiling and non-bolling conditions it is known that this value is overty conservative (Enclosure 2 to Reference 1 1) and that many factors will affect the G value. For instance, temperature has an effect on the rate of decompositiort When water is non-boiling, higher temperature usually reduces the radiolytic gas concentration. However, as the water approaches boiling, higher turbulence _ increases the gas production rate, and the G-value inc. eases. At the point of boiling, most dissolved gases in the water are stnpped out and a higher G value is appropriate. The presence of impurities. such as dissolved fission products, which come from post accident fuel failure, have a strong effect with respect to increasing the magnitude Of the G value.

3.1 - Temoerature and Turbulence Effects on Non-Boillna Watfg The ORNL (Ref. 31) data shows the hydrogen partial pressure against integrated dose rate for a number of expenmental cases. The experinients simulated various representative BWR core flow rates (100 gpm,1000 gpm and 10,000 gpm), and used different cover gas compositions (air or 5%O2 /95%N;) and temperatures (67 C,99 C and 129 C). The water was distilled so no impunty consideratior? are involved.

l The test data generally concluded that:

7

a. From 66* C to 97 C, the G value decreases with temperature. It tums around when temperature is increased to 125* C (stBI non-boulng under pressurized condition),

b. Initial G(H2 ) varies from 0.1 to 0.3 for SWR representative core flow rates from 100 gpm to 10,000 gpm. A higher pumping rate corresponds to more turbulence and thus less recombination.

LCP. GEN l

l _ _ _ ,

TR481 Rev. 0 Pcge 11 c, Radiolytic gas pressure reaches an equi!ibrium (G =0) in each case. At equilibrium, the dissolved hydrogen will recombiae with any oxygen molecules produced by radiolysis. The net production is zero.

The Staffs Safety Evaluation (Enc. 2 to Ref.1) states that for pure water (no iodine), it was determined experimentally that with no dissolved hydrogen and no boiling G(02 )=0.08. This conclusion appears to be based on ORNL Case No.11, which involved 95% N2 and 5% 02 gas over distilled water at 65' C and a flow rate corresponding to 10.000 gpm in a BWR (Fig. 3.1). G(q) becomes zero when the hydrogen's concentration reaches 2.5 cc/kg corresponding to an equilibrium partial pressure of 0.16 atm (Ref. 3-3). This was used as an argument that the G value should be significantly greater than zero for pure water under non-boiling conditions. However, the high equilibrium pressure is mainly caused by the high pumping rate during the experiment (15 crd/ min) corresponding to 10,000 gpm in a BWR under forced flow conditions. Higher turbulence removes free radicals faster and thus reduces recombination, For post accident BWR conditions, all pumps are tripped and a natural circulation condition is in effect in the core. The flow rate under these circumstances is closer to Case 10 of Reference 3.1 (see Fig. 3-2); from which'we can denve G(02 ) becoming zero at a hydrogen partial pressure of about 0.04 atm (4% hydrogen in containment) or a concentration in water of about 0.6 cc/kg (Ref. 34).

Assuming a degraded core condition with 5 times the 10CFR50.46 calculated metal water reaction (2.24% MWR), the initial hydrogen concentration in the Oyster Creek containment is calculated to be about 4% (Ref. 3-3). This partial pressure of hydrogen under non-boiling conditions was shown above to result in a G(02 )=0.0 at equilibrium. The ORNL data case 10 thus supports the NEDO assumption of G(02 )=0 for non-boiling if no iodine was assumed in the post-accident water.

1 LCP. GEN

o TR 081

. Rev.O Page 12 Since the presence of iodine in post accident reactor codant cannot be ignored, tne pure water G value data wWI not be used in the Oyster Creek plant specific calculations presented in Section 5 of this report.

32 Boilino BoHing strips dissdved gases out of the liquid phase so that the maximum decomposition will proceed. Equilibrium between the atmosphere and the liquid will not occur during boiling. It is conservatNely assumed that post accident boiling WWI last 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> in the Oyster Creek combustible gas concentration calculation (consistent with NEDO-22155). A G(02 )-0.225 wWI also be conservatNely assumed for this entire duration. Enclosure 2 to Reference i states that the maximum values of G(0:) for 5% MWR and 30% iodine release are between 0.19 and 0 22.

3.3 Imourttles The presence of impurrties such as post-accident fission product iodine in the reactor water may affect the decomposition rate. A static model is used by the NRC (Ref. 3 2) as fdlows:

C(N2 ) =Gg -

Kn (# }

2 2

G(0 3) =

Where Cw, G5= Initial G-value of hydrogen and OH radical, molecules /100 ev

[Hr ], [l] = molar concentration 'of H2 and i K. , K = rate constant of the reactions between OH and hydrogen and OH and iodine j LCP. GEN i

i

o TR 081

. Rev, O Page_13 For a small iodine concentration in the water (<10' gm mole /l) G(H 2) decreases very quickly to zero as the hydrogen coricentration in the water builds up. However, for a .tioderate iodine concentration ( [l] > 10' gm mole /l ), which corresponds to greater than 2% of the total core iodine being released to the water, the G value remains "sttive arv1 would increase the long term oxygen build up in the containment.

I LCP. GEN l

e l' a ijit r . 1-I. ( si< f J t. 1 xga s eti. ti t fJo. Ii Q4 ~~--~~-- -, __--.

. .se a r . seu 24 pqqq

~

q i#

C in ,8 < vr.s n s .. <, .

e w a , ora . u n . , , u . . . ,

o tA, lo i njijst, ,,a g i43, ,. ., ,

r., O 3 - --- ---

il MPt it Aloht

, g.,.(

8 u 9's t g c < e, t g i 60t 841) VOl eJPi . /UO 4 n.I

~

I iiILtItvi (JU*J HAli ,4 a 9ts

• sf su t.4 3 cr , iLO* H All , l's a m 3/m.n ( - to,ooi) .g,n. N ueu 3 3

eO2 -- ---

i A

+ -

. _ - . toyg g (,A.,, 9 27. N

• s7. O . I I A l M i

{ ,.

(,AS ANAt V S45 IN VUt OMt

", vi "f. Al i udse +1tmuu u

fg

+

[ A e

t e e,

=> 1 t o t, I9 tsd ..

In i a OI la - - e.

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

I ~ I- -

A u A e f  ; 7 ,,,, , se ss g / *a ./. t . o,,n. - ' < m-' ..u' e, o u i so o

is o

g It.l. n m f /[.

O[

O O$ tO t$ 2 O talO")

HAus A1 ON DO$l 18 )

.n x . 4 i's eD *n

<D < s CD

• O m

+* Q r*

Se

)

e l' A 4J u n - J./. Ul<fil. i:x;>+ t a ment f 441, iU s *** %s t*Mo 90 ' 144)2

~~-' ~ ~' ^- ' ' ~ - - '

' ' ~

O t$

It sl i t *4t ie t n 44 ,

P si N 'JJL til lufd, tJihllt i i t) ri gt) 2 GAS 10 t njue!) t<Al IO, J *. t

• It uet seAluHt A 42S*C g l

- . o 9n b C st%"L

~

t so nt) vot uun , i'Ou ... 5 g O to -- -- [ ~ i##t(levt t)(r,t Hall. 4 a t o' e/ w E 8 8

ItOW HAl t , O 15 (m /m.o ( 100 esp *** IN 81 AH )

tuvl H GAS, 9$7. Ng

• S% O,, I e A I M )

$ .f w l GAS ANAt VLtS IN VOLOMI % AI I U'Hi 'UI'"U O, te, g l r e, w

A .* A 3/ e, 9 u > ';

I go -

2 o e 36 66 *> 4 0 0$

i i lo . .g= a Al i t H h f/* f,

' A ts # oe<t gg A O lO" s> sS Sn $4 o e

/ # ## '

e , otsm ' ( m~ '

• 10' o O ~j $ *M, O O (f e l. s>pm N

o f , =.* *

• l a.

f ,~,, o ~

4 g j. . _ _ _ - _ . _

2 0 MA 0 i0 $

0 O$

HAut Al soft DOSI (8)

TR481

, Rev. 0 -

Page 16 l

40 OYSTER CREEK SPECIFIC IODINE RELEASE AND METAL-WATER REACTION 4.1 lodine Concentration from LOCA and LOCA with Deoraded Conditions Based on the NUREG/CR 2367 fission product release data (Ref. 41) an Oyster Creek specific iodine release concentration was cali .,!ated for both base case LOCA and LOCA with degraded core cooling. The degraded core cooling case is defined as having a metal water reaction (and resulting hydrogen generation) that is 5 times the amount calculated for the base case LOCA used in this evaluation. The metal water reaction rate used in the base case LOCA was somewhat greater than those calculated pursuant to 50.46(b)(3) because of the ECCS code used in the evaluation. The released iodine concentrations are calculated in this section to be 1.80E49 and 5.44E47 gm mol/l for the base case LOCA and the LOCA with degraded conditions, respectively.

The Oyster Creek core wide metal water reaction based on compliance with 10CFR50.46 is 0.448%

The 10CFR50.44 criterion for degraded core conditions is the larger of: 1) five times the amount of hydrogen calculated in demonstrating compliance with 10CFR50,46, or 2) for the amount that would result from reaction of all the metal in the outside surfaces of the cladding cylinders surrounding the fuel to a depth of 0.00023 inch. A fivefold increase in hydrogen corresponds to a fivefold increase in the metal water reaction which would be 2.24% The metal water reaction due to the reaction of C 00023 inches of the cladding surfaces is 0.77% Therefore, the fNe time increase in MWR is the enterion used for Oyster Creek in determining the degraded core condition.

A fuel heet up calculation was performed to determine fuel rod temperatures, MWR and the number of faaed fuel rods during a LOCA based on an end of cycle (EOC) core conditions. Using the NUREG/CR 2367 iodine release rate, the total. lodine concentration was calculated along with a core

j. wide metal water reaction. This served as a basis from which the degraded core cooling case could be evaluated. For the degraded core cooling case, it was assumed that the initiation of emergency ccre cooling was delayed and flow ratos were less than Appendix K requirements. The fuel heat up r

LCP. GEN

TR 081 Rev.O Page 17 calculations were redone with reduced ECCS flow and iterating on the time for delayed core cooling until the metal water reaction increased by a factor of S. An iodine concentration was calculated for the degraded core cooling case using the resulting fuel rod temperatures and Reference 4-1 fission product release data.

l i

The heat up calculations were performed using the HUyY code (Ref. 4 2). The HUXY code has i been approved to perform 10CFR50 Appendix K calculations for the ANF fuelloaded in Oyster Creek. The HUXY code does not calculate the mechanica' response of the cladding during a LOCA.

However, it does allow a temperature input which, when erceeded, fails the fuel rod and calculates a MWR based on both an inner and outer cladding surface at per Appendix K. Current licensing analyses (Ref. 4-3), show that a fuel rod will perforate at nocal exposures exceeding 19.0 GWD/MT when the peak clad temperature (PCT) exceeds 1600 F. Disiding by an approximate axial exposure peaking factor of 1.25, this translates to a bundle average exposure of 15.2 GWD/MT. For bundles having exposures less than 15.2 GWD/MT, a fuel failure temperature of 2500 F was used.

i i

A core power distribution histogram (Figure 4.1) of number of bundles versus radial power was developed for bundle exposures below and above 15.2 GWD/MT. An EOC case was used for conservatism to maxim'2e the number of higher exposed fuel bundles. Another conservatism was to l

place all fuel bundles in the peak radial power group, for the high and low exposures, at their MAPLHGR limit. The heat up calculations were repeated for the low and high exposures, for each of the radial power factors indicated, and for the base case LOCA and LOCA with degraded core i

cooling conditions.

l

! The results of the calculations are summarized in Table 4.1. The base case LOCA calculations result 1

i in 6288 failed fuel rods out of the 33,600 rods in the core, and a core wide MWR of 1.16%. Both of these values are greater than the Appendix K results due to the additional conservatisms used in this l

LCP. GEN

'$ h o

TR 081 Rev. O Page 18 analysis. The degraded cooling case results in 17744 failed fuel rods and a core wide MWR of 5.85% (an increase of a factor of 5.04 over the base case LOCA case): The iodine release rate was calculated for each group of bundles for a given radial power factor based on the calculated fuel -

temperature. A fuel rod lodine concentration of 0.486 gms per fuel rod, which corresponds to a high power rod, was conservatively used for all failed fuel rods in the core. The average core lodine concentration is 0.389 g/ fuel rod (Ref. 4-4). In addition, the lodine release rate for a failed fuel rod was conservatively based on the limiting (hottest) axial node in the rod.

4.2 Relationshio Between Metal-Water Reaction and lodine Release The analysis discussed in Section 4.1 provides an estimate of the core wide MWR and lodine release for degraded core conditions. The treatment of the MWR was based on the parabolic rate law of Baker and Just and the iodine release was determined using NUREG/CR-2367. Figure 4 2 is a plot of the MWR and iodine release as a function of pin centerline temperature for a failed fuel pin. As can be seen, if the pin centerline temperature increases, both percent MWR and lodine release also j increase. As the temperature increases to 1600* C, the iodine release approaches 30% while the l'

MWR approaches 70% of total.

j 1600' C represents a limiting fuel pin condition for the degraded core analysis reported in Section 4.1. While'a few pins may approach this limiting condition, the majority of the fuel will remain well below this temperature. The inset in the upper left comer of Figure 4 2 lists the degraded core analysis results for percent of total core MWR and percent of total core iodine release. The lodine release and MWR percentages reflect the fact that for the degraded core condition, only half the pins fail and the centertine temperatures of most of the pins are well below l 1600' C.

- LCP. GEN

• M*<0.

Fev. O l

1 Page 13 1

i . . . . I g  % g g  % s g J

e s s's J e ' s x? n bN'-Js'.?q'.N'?x00-a>Nca pmM' .sn\x,N'Jq  ? - CO o

i

! O'

. . O E i  :

i h "" ' e cn.

r_.

CO e--

l  :  : o y Q ,

q O .

O m

.E,

_-- l l Nk.h'N

% mm A

o o c._

b i

i *

: uo=W m.

C

' [

C W

• • g

- ^

(O

. .  : L1_ s .

L C

l .

s mi. 1 i

9. A

-' l

- -( .

F-

-CD+CO .

y e-- p .

to a m

~

I C

'  ! a'mM k 'kh e s' h g &

h'NxNQ'.NN z 3 l i

sk%39E" O m

& en

i. i e

O Q ~

! {

I

\

O  ; - '

D 1 O  :

g . . .

.c Q I

' f  ; o xxx1 x v11 x=

T

L4E.wM,v a. w t4 a  : <

Z i i i i i o o o o o o o N O CO LO v N w w

_ . _ . . . . . =m , m. _ .._ . . _ . _ . _

. -o

-TR-Cal Rev. 3 Page 20 5

$4N O O O O O O + O O+ O +m o+

@ O e

'O

• O O O O

ei O

8 4 I i + +

WWWWW W W W W W W g

mI Q

O O O O O O O O O b b b b n

w M

O e

m N

O O

O O

O e

N E b b N u

y e e 6 m m N m to w W m rw - M m N - M o O E W O O O O O O O O O O e= b o O O O O O O O O O Q at O O O O O O O O O O m b b b b b b d b b b I

"8 0

8 m m N w M N O @ m N

{ A E b k 0 b b b l a

b 0

2 A O O O O O O O O O m 43 O m - @ @ 4 @ m @ m a N O O @ m o @ @ m @

3 m m m m m E

S b $

3 hm he O O O O w N

w m

O o

O e

O O $

N O mO @ m N m e at at M "

e 6.

e e

e

==

9 K M N

5 4 m N O N N O O

• N O O I

& v m N m M w m m C w @

3 m

g O

O

.ar 9 b l 9 8

! b U C 7

g% W l

q e m C O O O N N O Q O p y A m m M

m $

. u

> , a e m

O

• W X B m

O h e O m w m N eg M N m M -M C "

N N N m b b *

• um 88

. ,It, n

.s N.

e N.

m N.

m N.

e N.

e N.

e N.

e N.

m N.

N.

O m m eg ce M m m og m eg

& V V V V A A A A A A M

M

__ . . _ . _ - . , _ _ _ _ . _ _ _ _ _ . - - - . _ _ _. - . . - . , _ _ .., _ , , . . ~ . . .

O TABLE 4.2

' Oyster Creek Iodine Release During LOCA with Degraded Core Cooling EXPOSURE RADIAL FAILED BUNDLES TOTAL. C/L FUEL MWR POWER IOLINE IODINE OWD/3rr ' B006/ FAILED TEMP C RFTFASE CONCENTRATIOtt m RODS ,'

RATE ON MOL/L

<15.2 1.48 15 76 1140 1460 16.3 0.067 1.10E <1522 1.3 0 48 0 1390 7.8 0.037 0.OOE+00

<15.2 1.2 0 32 0 1310 5.8 0.0023 0.OOE+00

<15.2 1.1 e 20 0 1210 3.7 0.00053 0.OOE+00

>15.2 1.3 60 12 720 1510 26.9 G..) 1.56E-07 i

>15.2 1.2 60 112 6720 13AO 17.5 0.028 2.71E-07

>15.2 1.1 60 40 2400 1260 10.2 0.0014 4.64E-09  :

>15.2 1 60 80 4800 1140 5.3 0.00029 2.01E-09

>15.2 0.9 31 32 992 960 1.8 0.00014 2.OOE-10

>15.2 0.8 9 108 972 910 , , 0.7 0.00008 1.12E-10 ,

N ALs: 560 17744 5.44E-07 .

l '

a 1

TOTAL CORE t IODINE RELEASF.D = 1.4 i

TOTAL CORE % MWR = 5.85 w" E N '

1. .< a >

u*l 1

e Rev. 2 FIGURE 4-2 ne 22 FAILED PIN MWR AND IODINF PELEASE AS A FUNCTION OF TF 'T N ' - I J R E PERCENT OF TOTAL 80 DEGRADED CORE RESULTS Percent of Total Core MWR lodine Release 60 - s.s s 14 i

40 -  ;

/

f 20 - ,

l

/

/

/

0 '

O 500 1000 1500 2000 PIN CENTERLINE TEMPERATURE (deg C)

=

% MWR .% lodine Release

$ TR481

.- Rev./

Pago,J3 50 PLANT SPECIFIC OXYGEN CONCENTRATION WITHffDl@ED G VALUEJ 5.1 Pethodoloav The Oyster Creek plant specific oxygen concentration was calculated for a variety of iodine and MWR assumptions. The methodology described in Appendix A of the NRC Standard Review Plan (Section 6 2.5), NUREG 0000

• Combustible Gas Control in Containment' (Ref. 51), was used except t

that G values are calculated as a function of dissolved lodino and hydrogen (Ref. 3 2) (see Section 3.3):

G(Hg )=0.45 2.7/(1 + kl [1]/kh (H]) (5.1) where G(H,) - not hydrogen generation rate, molecules /100 ev

[1]

= lodino molar concentration in the coolant

[H] . hydrogen molar concentration in the coolant kl,kh - reaction rato constants for the adverse lodino reaction and the favorable hydrogen reaction in radiolysis suppression >

For the first 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> of the LOCA, G(H,) is given its maximum value (0.45), as in boiling Thoroafter, G(Hg ) is calculated from Equation 5.1. The dissolved hydrogen is calculated from Henry's Law; PH2 = KH'[H] (52i i

where KH = the Henry Law constant for hydrogen in water

PH2 = the pressure of hydrogen in the gas phase LCP. GEN

Ow 0  !

Poge ;4 i

The detals of the application of these equations to the speerfics of the OCNGS are gNen in Apperdix A.

52 Resu'13 5 2.1 .Qyiter Creek Soecffic Orycen Concentention The results of the HUXY analyses discussed in Section 4.0 showed that for a br.se case LOCA the iodine concentration will be 1.80E-9 g-moles / liter which corresponds to a total core iodine release of abou' O 0046%. The total core % MWR for that case was 1.16% The containment oxygen concentration for this event would only increase by atnut 0 25% as can be seen in Figure 51.

For the degraded ECCS case analyzed, the total core iodine release was 1.4% and the calculated % MWR was 5 85% (a factor of 5.04 increase over the base case LOCA case).

The orygen coacentration in containment for this cace would not increase from the initial value. This is depicted in Figure 5 2. Figure 5 3 provides the results for the same iodine release case (14%) with a MWR of 2.24% (5 times the 0.448% calculated for 10CFR$0.46).

Again, a 5% containment orygen concentration is not reached.

52.2 Oxvoen Concentration Followino Severe Accidents 5.2.2.1 Total Core in this section, iodine and MWR assumptions more severe than those calculated

! specmcally for Oyster Creek in Section 4.0 are evaluated with respect to expected oxygen concentrations. Thesa analyses are being performed to address the release of larger amounts of iodine uh to and including 30% of the total core iodine inventory. The release of such large fractions of the total core iodine inventory would require that all of the core fuel rods achieve substantial fuel center!ine temperatures (Ref. 41). Fuel rods achieAng such high centerilne temperatures

(

LCP. GEN i

_ _ __. -.~.

Rev. 0

. P0ge 25 would also be undergoing substantial metal water reaction. The re'ationship between these parameters was discussed in Section 4 2 of this report and will form the basis of the cases analyzed herein.

Figures 5.4 through 5 6 show the oxygen concentration profiles for 10% 20% and 30% total core iodine release The 10% iodine analysis (Fig. 5 4) shows oxygen concentration for a % MWR of 15% The results indicate that it will take about a year to reach 5% oxygen concentration. Figure 5.5 shows the 20% iodine release results with a % MvVR of 30% Again, oxygen concentrations of 5% do not result in less than approximately one year. Similar results can be seen in Figure 5 6 fo'r the 30% iodine release and a % MWR of 40%

The selection of the % lodine /% MWR ratios was based upon the results depicted ln '

Figure 4 2 which shows that for a gNen fuel temperature condition that results in a

% lodine release, the % MWR for that condition will be substantially higher than the )

corresponding % iodine. A conservative ratio of 1.5 or less was used for each case analyzed.

5.2.2.2 Localized Effects This section is addressing the concem that, in the event of a LOCA, a small fraction of the core might become overheated, it is assumed that this might occur from a hot spot resulting from local coolant flow starvation as a result of: 1) delNery of less 1

than planned cooling to a localtzed area, or 2) local l'ow blockage. It is further assumed that the expected hWR will not occur at any time even though such an assumption is not credible.

j. The following conservative assumptions are being used:

LCP. GEN I_...,_-.-- . . - - . . . - . . - - - - . . - - - - - . . - - - - ~ ~ - - - - - - - - - - - - ~ ~ ~ - ' ~ ~ ~'

- . _ . . . . - . . . . . - . . . . . = - _ ~ - - ... _~.. - .. -.

. Rev.0

, Pege 26 a) Fuel center 1ine temperatures in affected region reach 1600' C (30% iodine telease).

b) Size of affected region is 10% of core (56 bundles).

c)  % MWR for affected region is 1%*

d) Remainder of core as per degraded LQCA case of Section 4 i.

This is conservatNe since higher percent MWR would produce less oxygen than the assumed case because of the suppressing effects of hydrogen on G(0:). The percent MWR would be about 70% for the affected region.

These assumptions result in an overall total core iodine inventory release of:

(1.4%) * (0.9) + (30%) * (0.1) = 4.26% I The total core % MWR is:

(5.85%) * (0 9) + (1.0%) * (0.1) s 5.36%

i The oxygen profde resulting from this condition is shown in Figure 5.7. The results show that a 5% oxygen conctatration in containment for this non credible assumption will not be reached for approximately three months.

I l

5.2.2.3 lodine Release WMbout MWR ,

lodine release without a comparable MWR is not credibis. Even if a blockage of cooling water to a small region of the core is assumed as the basis for limiting the MWR, eventually either cooling wel occur or fuel melt will result Melt progression l

1 wel only cease when cooling'is re estabilshed. When this occua, MWR Mi also occur. The requirement of 50.44(h)(1) is that the degree ci degradation is not  ;

sufficient to cause core meltdown. This implies that cooling is establ3hcJ, and this cooling of hot fuel must result in signMcant MWR.

LCP. GEN

_ _ , . _ _ . . _ . . _ . _ ___ _ _ _ .- __ _ . ~ _ _ - - _ , ~ . . _ _ _ _ _ _ . , . - _ . . _ _ _ _ _ . _ . . - _ . _ _ _ . . . _ . . _ . . _ _ . _ _ . _ _

o Rev 0

, Page 27 At TMl 2 where cooling was unavailable such that significant core heatup occurred, a significant MWR resulted from the eventual re-establishment of cooling water.

Even rf complete core mett were to occur. significant MWR would occur when the melt material comes in contact with water inside containment.

5 2.3 Addttional Conservatitim3 a) The NRC model used in this report (Section 5.1 and Appendix A) is very conservative (overpredicts G value) when it is applied to low impurtty (low iodine and hydrogen concentrations) cases. The reason for this is the assumption of an inrtial G(%) of 0.45 which is then allowed to decrease. In the Zlttel experiment for pure water (Ref. 31), the G value never exceeds 0.3. This dtfference contributes to a larger radiolytic gas production rate and higher oxygen concentration in containment. For iodine concentrations less than approximately 10 4g. moles / liter, it would be more appropriate to apply the Zittel results.

b) The calculation herein assumes that the precipitated ZrO; from the metal water reaction would occlude 10% of the water and that this water would have a G(H2 )=0.45. The NRC in Reference 3 2 uses a value of 1% rather than 10% for this effect. The model herein would thus overpredict the radiolytic gas production slightly as a result of this.

LCP. GEN

I l

t FIGURE 5-1 BASE CASE LOCA 4.000 1

3.942 n -

6 3.884 '

l z -

9 3.826

>h3.768 x x x x d x x ;cx xxxx xxxxx xxxxxxxxxxxxm z - 9xxx d 3.710 1 z -

0 o 3.652 .

z w 3.594 o .

l I

N 3.536 l o 4 '

3.478

3. 42 0 360 0 45 90 1 3." -' " ' 1' ' '8 0 225 270 31 5 TIME (DAYS)

/

na

'e - at tJ~b a

i t

a i

4 FIGURE 5-2 i IODINE =1.4%: MWR=5.857.

l 3.60 ,

i - t t

3.57 ,

n -

5 3.54 - -

l

z -

I 9 3.51 i s'

! l l {g 3.48 ,

! z :xxxx x x x x :. x x x x x x x x :: x x x x x x x x ::xxxxxxxx  !

O 3.45 i .

z - '

i O o 3.42 z i i to 3.39  :

O _

N 3.36

, O _

1 l

3.33

_ t 3,30 ........ .. .....

l 0 45 90 135 180 225 270 31 5 360 TIME (DAYS) i  !

mxa

. =ex <

Y 7J <1

'D - O [

' oc e

$O M W

, c l

.i 4

k i

l l <

i

. i e i

i FIGURE S-3  :

1 IODINE =1.4%: MWR=2.247. .

1 r i

4.50  !

, - p' r-~

i l ] ,

! 4.38 .

5

+--- l l e

- 4, ,

m .- I  : , i -

i l-i- 6 4.27 l l k- *

-- - i  !

! z -

i , f '

l  !

j

! 9 4.15 '

L---- 2

l. l 1 H l t

' l I I '

s 4 .04 4

.  ; i  ;

1' i b

O 3 92

""*******E**** *************E j 4 g j

z -

I . l  :

d 4

l i l 83.81 ;f  ; 1 l z I i r i l l l w 3.69 i. i

~

4 -

4 e i i

1 '

k- 3.58 _.,I -

J i

! o / I

- 3.46' i --~ l ,

e  ;

I '

3.35'"'""'""""""""'"""'"'""'" I l 0' 45 90 135 180 225 270 31 5 360 i i TIME (DAYS) '

3 i

,me  :

__ _ e j$7 [

e- o  :

< m i

, WW W o -

3 I i . .- .

i

l.

1-s i

\

i ,

FIGURE 5--4 l IODINE =10%: MWR=15%

! l 5.00

g. g er. . . - - ;

x*e: W 4 4.77  :

• 6 4.54 z

j*#1 p l '

--b

!  ! i l

i 9 4.31

?p ---

i' s

r i

h r 4 .08 e ,

e

. z -

/ i i t' i d 3.85 l' , -! i z -

i o 3.62 o

i -

I t i w 3.39 E ]/

2 .

l I  !

! -j i-

o l

l r '

I l D3.16 7

' l i

o b e

i 2.93 -

l l 1 2.70'''''''''''''''''''''' ' ' ' ' ' ''

O 45 90 135 180 225 270 31 5 360 i' TIME (DAYS) t

, --J =c -4

~ ~ - ~ ~ * ~

7 ,f,

u-S r . - -

1 i .

c

i l

l RGURE5-5

IODINE =20%; MWR=30%

. 5.000 .

4.750 --

o -

.,m. , ,

6 4.500 ~ ' ' '

z 0 4.250 x

W , x :' l l

l

! 4 000

1. xY 5

$ 3.750 y f l

. z -

t

!' O o 3.500  : .--

i z

w 3.250 -

t

e -

! D 3.000 i i

0 2.750 -[  ;

/  ;

2.500"''''''''''''''''"''''''''''''''''

O 45 90 135 180 225 270 31 5 360 t

TIME (DAYS)

=nm

.,x .

.

• O ,

g3  ?

4 t

t

l -

i 1

I l FIGURE 5-6 IODINE =30%: MWR=407.

I i

4.999 --

l 4*695 p W ' ' ' "

, v 4.391 -

z -

i 9 4.087 #

1 y

l h.783 s 3 p I z -

d 3.480 -

f z - -

4 03.176 O

z

w 2.872 -

o D 2.568 2.264 _l[

i O -

l o f

1.960"""'"""""""'""""'""-""" '

O 45 90 135 180 225 270 31 5 360 I TIME (DAYS) ,

i S~~

\ ,

l

lj!ll lll l lll' ll!i 1 l me3

? ~

n% W l

• i q .

0 _

- 8

- = '

j;!I'tl 0

- ' 7 g -i 0

- 6 6 "

3 -

5 '

0

= ' 5 7R -

)

'[

W S 5M ' Y

" A E :. -

0D R 67 "

4(

U E G2 j -

" M R4= -

I T

E g -

0 3

,/

N I "

D -

O -

I g '

0 2

/

0 3

1

/

/ '

/ '

/

- - - - - _ - _/ c. -

0 0 0 0 0 0 0 0 0 0 0 0 0 8^ 6 4 2 0 8 6 4 2 0 5 4 4 4 4 4 -

3 3 3 3 3 gv z$gwzdzO0 zwoNo i ,  : ,liilj! ' l;; 1i;';!l:i: :i i iij: !f !:  :

, m vo i Rev.0 Pag 3 35 l

l TABLE 5t SUMM ARY OF RESULTS l

CASE DESCRIPTION  % IODINE  % MWR TIME TO 5% OXYGEN (DAYS) l Base case LOCA 0.0046 1.16 > 1000 LOCA wah Degraded ECCS 14 5 85 > 1000 LOCA wah Degraded ECCS 14 2.24 > 1000 Severe Accdont 10 15 360 Severe Accdent 20 30 > 1000 Severe Accdent 30 40 > 1000 l

Localized Effect 4.26 5.36 90 l

I LCP.CEN

' .n w AW O ,

Pc9) 36 l

I 60 CONCLUSIONS The following conclusions can be reached as a result of the analyses discussed in the report:

a) The non boiling G(q) is not zero with the presence of dissolved iodine, but the effect of the dissolved lodine in the containment water on the G(02 ) is offset by the effect of the dissolved hydrogen resulting from the initial metal water reaction and from radiolysis.

b) For both a LOCA and a LOCA with severely degraded ECCS performance, the oxygen concentration in the Oyster Creek containment would not reach 5%.

c) For severe accidents in which 30% of the total core iodine is released, i.e.. NRC assumption of fuel rod centertine temperatures of 1600' C over the entire core, the oxygen concentration would not e reach 5% in less than approximately one year for a conservatively low zirconiurr, water reaction rate of 40%. For zirconium water reaction rates greater than 40%. It would take even longer to reach 5%.

d) There is no credible mechanism by which substantial amounts of core lodine can be released without a substantial amount of metal water reaction.

e) Even in the event that flow is blocked to a small fraction of the core following a LOCA, crygen concentration in containment would not reach f % for several months.

LCP. GEN-

.___ _ _ ____. _ _ _ . _ - . . _ _ _ _ _.._..__. - ____ m.___ _ _ - m _

,nw, i Rev. 0 l Pege p l l

70 REFERENCES

.11 Steven A. Varga, USNRC Letter to R. L Long, GPUN, November 6.1990.

12 BWR Owners Group Report.

• Generation and Mitigation of Combustible Gas Mixtures [n inerted Mark i Containments", GE NEDO 221511982.

31 ORNL TM 2412 Park Vill.

• Design Considerations of Reactor Containment Spray Systems . Part Vill i

Boiling Water Reactor RaJidysis Studies *, H. E. Zittel, October 1970.

l 32 Menv from K. I, Parczewski to Victor Benaroya, *Radidysis of Coolant Water in Mills:one l', i June 23,1982. i 33 GPUN Calculation C1302 243 5450460, 'OC Post LOCA Hydrogen and lodine Concentration h the  !

Containment and GayProduction Rate by Radidysis', June 1991.

.41 ' Updated Best Estimate LOCA Radiation Signature,' NUREG CR 2367, D. D. Thayer.

.l 42 HUXY: A Ger'eralized Multipod Heatup Code with Appendix K Heatup Option XN CC 33(A)

July 28,1975. .

43 ' Oyster Creek NGS SAFER /CORECOOL/GESTR LOCA Loss of Codant Accident A.Wysis,' August 4

1987.

4 -4 GPUN Calculation C1302 226 5411-236,

• lodine Release During LOCA wtth Degraded Core Coding *,

May 1991.

51 NUREG 0800. USNRC Standard Review Plan, Section 6.2.5, ' Combustible Gas Control in Containment *, Appendk A. U

'A1 NUREG4800, USNRC Standard Review Plan: Section 6 2.5 Combustible Gas Control.

l A2 US NRC Womorandum, Kl. Parczewski to Victor Benaroya, Chief, Chem. Engrg Branch, Div. of l Energy, *Radiofyse of Codant Water in Milstone 1.  !

i l A3 GPUN Calculatkan C1302 240 6340 005, *Radiolysis After LOCA at OCNGS*, May 1991.

1 i

LCP. GEN l

...,,.w.,,.......,_..,, . . _ . . _ . _ _ _ . _ _ , _ . _ _ . _ _ _ . _ . _ . . _ . . . _ _ , . _ . . . . . . _ . , _ , _ _ _ . - . _ . . _ . _ . _ , . _ , . . _ . -

Ret 0 Page 38 APPO4DtX A OXYGEN VS. TIME CALCULATION METHODOLOGY INTAODUCTION The SRP (Ret 1) Section 6 2.5 calculations are used wah modifications as r oted.

OtSCUSSION in the event of a lossof<oolant acedent (LOCA), hydrogen and oxygen gases will be generated wnhin the Creek reactor containment by:

1.

Metal water reaction involving the 2irconium fuel cladding and the reactor coolant. producing fre 2.

Radiolytic decomposition of the post accident emergency coding solutions, producing both oxygen a hydrogen.

1 if a sufficient amount of hydrogen is generated, it may react with the 0, present in the containment atmosphe

• in the case of inerted contain e tm n s, with the oxygen generated following a LOCA.

The extent of 2irconium-water reaction and associated hydrogen production depends strongly upon the course of events assumed for the accident. Analytically the reaction can be described by:

1 Zr . 2H 2O - Zro, + 2H, 1 lb Zr - 0 043956 lb H,

- 1 lb Zr - 0.021978 It> mole H,

.1 Therefore, one pound of reacted zirconium w#l produce 0.021978 pound moles of free hydrogen. Assuming the i perfect gas relationship, this is equkalent to 8.4866 scf/lb Zr:

v=yn '

V = 0.021978(10.71)(530) /14.7 (Standard conditions taken as t4.7 pala,530' R)(70' F)

V = 8.4866 scf/lb Zr.

LCP.CEN L

f

._+w-e, ,e -,r-,. +r.--++v *<_w-a,. ere- ,.,-...,e. m--,-e-r-.- _m-ww

-- ,, -e r-w_.e. , _r.+,- ..c+.e e y ,w ~ w.r w w -- y- et.,ax-m--g-g-wyt---m

. TR 081

, Rev./

Pagej'[

The total amount of hydrogen produced is based on the amount of reacted zirconlum. The computer program, to maintain a degree of generality, a! lows the reaction percentage to be specified as an input quantity. The expression used is:

WG = (.022)(WZr)(t,,,,)

where WG = pound mdes of hydrogen generated WZr = weight of zirconium fuel element clad tu = zirconium-water reaction fraction The arte of gas production from radiolysis depends upon the power decay profile and the amount of fission products released to the coolant. The radiolytic hydrogen production rate at time (t) is given by.

S y (c)- = (EC( t) + E,( t) )

(B) (N) 100 ( B) (N) 100 where

((t) = hydrogen production rate,Ib-mole /sec P = operating reactor power level, MWt

~

B = conversion factor,454 gm-mde/lb-mole N = Avogadro's number 6.023 x 10" molecules /gm-mole Q = radiolytic hydrogen yield in core, molecules /100 ev

((t) = gamma ray fission product energy absorbed by core coolant, ev/sec-MWt Q = radiolytic hydrogen yield in solution, molecules /100 ev See below for definition of G(H2 )

((t) = energy absorbed in coolant outside core due to fission products dissolved in coolant, ev/sec-MWt The quantity ((t)is defined by:

((t) = (Q F4(t)

LCP.CCN

-I w.

• Rev.O l Page ac

)

l where (f, L; = fraction of fission product gamma energy absorbed by coolant in core region

= 01 H,(t) = gamma energy production rate, ev/(sec-MWt)

Simdarty. ((t) is defined by:

((t) = (f, . , k H, . ,(t) + t H (t) where (t,. ,), = fraction of total solid fission product energy absorbed in coolant outside cors

= 0 01 H,, ,(t) = total sold fission product energy production rate ev/sec MWt

( = fraction of iodine isotope energy absorbed in coolant outside core

= 100% of the fraction of lodine energy released to the coolant H(t) = iodine isotope energy production rate, ev/sec MWt The equations for oxygen generation by radiolysis are identical to those above describing hydrogen evolution except that the yield is one-half that of hydrogen. For calculational purposes, the reactor decay profiles (H,(t). H, . ,(t),

and M(t)) specified by the ANS 5.1 standard for two year reactor operation have been fitted by several finite exponential series expressions and also incorporated into the program. The resulting equations are:

N, ( c) = 1088 ( 5 .1912 e '"'' +0. 87 4 3 ed '" ' + 0 . 6 5 57 e -' '"'" '+ 0 . 4 0 9 8 e -' * "' ' ' + . 010 5 0 e '"'"* *

1. ,., ( c) = 2 . ON, ( c) l l

1. f(c)=108 8 ( 0. 819 7 e ' 8"' + . 3 2 7 9 e -1 8 3"' + . 0 57 41 e *"' '

I l

l l

l LCP. GEN

• TR481 Rev./

Page 41 where t = time after reactor shutdown, sec.

Between 400 and 4 x 10' sec, the equations overpredict the standard curve by 20% The equations underpredict the -

standard curve soon after shutdown. However, this does not seriously affect the results due to the short time period involved. The equations are equivalent to the afterheat decay curve in BTP ASB 9 2 over the times of interest for post-accident hydrogen generation. It should also be noted that this formulation overpredicts the radiolytic hydrogen generation by a small amount due to a " double counting" of the gamma energy of those fission products assumed to be released from the fuel rods.

G(Hg ) is taken as

= 0.45 during bolling

= 0.45 - 2.7 (1-fz)/(1 + ki'[l]/kh2*[H]) in non boiling water (A1) where ki = le10 l/(m*s) (liter /mol sec) kh2 = 3e71/(m*s)

[l] = dissolved lodine, mol/ liter n [H] = dissolved hydrogen, mol/ liter i

' fz = aee below l

We assume (as per A. O. Allen, Ref. A 2)

1. Water system consists of suspended ZrO2 and dissolved iodine in water.

2. Water included in the porous ZrO2 particles continues to boil.

3. The fraction of water (and decay energy) absorbed in the ZrO2 is fz, and this is the same in both the core and torus (fz=0.10).

LCP. GEN

• . n vo i Rev.0 Page 42 The dissdved hydrogen at any time is calculated from

[H] = Ph2iKH (A2) where Ph2 = partal pressure of hydrogen in the total gas vdume, (wetwell + drywell), psia KH = Henry's Law constant psia /(md/l)

A basic computer program performs these calculations The hydrogen inventory is calculated by step wise int)gration of the hydrogen production rate, the Ph2 is caiculated assuming a perfect gas in the drywell, the [H]

calculated from Equation A 2 and the G value for the next time increment is calculated from this [H] value and A.2 to repeat the cycle.

The percent oxygen at each step is calculated from

% oxy = 100% mox/mox + mh+ mn) where mox = total mdes oxygen (original inventory plus 0.5 times the mds hydrogen produced radiolytically) mh = total moles hydrogen (radidytic plus Zr/ water reaction mds) mn = total mdes nitrogen originally present The radidytic hydrogen formed during boiling (the first 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> LOCA) can be calculated analytically since the G is independent of time, and the decay energy expression integrates to a sum of terms in the form B'(1-EXP( C*t)) with b and c constant and t= 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> I

j LCP. GEN i

, i n vo i

.* RW. 0 Pege 43 APPENDIX B SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION GENERAL ELECTRIC COMPANY'S METHODOLOGY FOR DETERMINING RATES OF GENERATION OF OXYGEN BY RADIOLYTIC DECOMPOSITIQHEEWng)

LCP. GEN l ._ - - ' ~

ie

[po sseg\ UNITED STATES

[,3#c g NUCLE AR REGULATORY COMMISSION

, j was m 0To v o c 20sts

(, v /

• ...e ENCLOSURE 2 SAFETY EVALUATION BY lHE OFFICE OF NUCLEAR REACTOR REGULATION GENERAi ELECTRIC CCMPANY'S METHODOLOGY FOR DETERMINING RATES OF GENERATIONS OF OXYGEN BY RADIOLOYTIC OECOMPOSITION (NE00-22155)

In June 1982 General Electric (GE) issued the subject report containing i description of the methodology for determining rates of generation of oxygen by radiolytic decomposition of water in the inerted Mark I containments. In this report, GE assumes that after an accident water in the containment will boil for 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> only. During this time it will undergo radiolytic decomposition with oxygen generated at the rates corresponding to G(0 3)=0.1. Where G(0 2 ) is a number of molecules of oxygen generated by 100 ev of radiant energy absorbed.

This value was based on the results from the measurements of the hydrogen evolution rate in the offgas systems during normal (boiling) operation and during refueling shutdowns and confirmed by the experiments performed in the KRB Nuclear Power Plant.

For radiolysis of water beyond 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />, when boiling ceases, G(0 )=0 3 was assumed and consequently there was no net generator or radiolytic oxygen.

This last assumption was based on the analytical results obtained by Knolls Atomic Power Laboratory (Reference 1) and by Argonne National Laboratory (Reference 2) in connection with the Three Mile Island accident. The values of G(03 ) in the GE report differ considerably from the value of G(0 )3 in Regulatory Guide 1.7 which for both boiling and non-boiling cases recommends G(0;)=0.25. However, this value is not based on any specific mechanism of rad 1olysis but is chosen to bound all possible cases and consequently it tends to overpredict the rates of generation of radiolytic oxygen. In 1982 an extensive effort was undertaken by the Northeast Utilities and by the NRC in connection with the Millstone 1 licensing action to determine a more realistic method for calculating rates of radiolytic oxygen generation, in performing this task the staff was assisted by a consultant from BNL. The results of this effort: have indicated that G(0 )3 is not a constant parameter but varies with the amount of hydrogen dissolved in water and with the concentrations of certain impurities, most notable among thee iodine. Since concentrations of these substances may vary with time and may be different for different accidents, the true value G(03 ) should be expressed as a function of these variables.

In general, an increase of ccncentration of hydrogen in water results in a decrease of radiolysis due to promotion of recombination reactions. On the other hand an increase of iodine concentration tends to promote radiolysis by destroying free radicals which are required for the recombination reactions to proceed. The highest rate of oxygen generation is achieved when G(03 )=0.22, 1

il() [ h. -.

]p

4 TR-CSI Rev. 0 2 hge 4$

which is the highest theoretical limit for gamma radiation. This occurs when

-ater is completely free of dissolved hydrogen, or when the concentrations of dissolved iodine are entremely high. However, in most cases G(0 2

) *ill D' lower and at certain concentrations of hydrogen and iodine the rates of radiolytic dissociation and recombinations reactions may become equal resulting in G(0 2)=0 and no net generation of radiolytic oxygen. During the boiling regime hydrogen will be stripped by vapor bubbles and it is expected that G(02 ) will be higher than in non-boiling water.

Quantitative evaluation performed by the staff was based on the model developed by the BNL consultant (Reference 3) and on the experimental data from ORNL (Reference 4). For pure water (no iodine) it was determined experimentally that with no dissolved hydrogen and no boiling G(0 )=0.08. 2 However, when under non-boiling conditions the concentration of dissolved hydrogen reached 2.5 cc/kg of water, corresponding to equilibrium hydrogen pressure of 0.16 atm.,

G(02 ) became zero and generation of radiolytic oxygen stops. This finding contradicts the information in the GE report where G(0 )=0 2 was assumed for all non-boiling cases.

For water containing dissolved iodine no applicable experimental data were available and the staff calculated G(0 )3 corresponding to the maximum credible iodine concentration in water using the BNL model. Since all iodine in t.1e containment water comes from failed fuel, an accident had to be postulated which would result in a release of this amount of iodine. In such an accident fuel was assumed to fail by oxidation of Zirconium cladding and hence, in addition to released iodine, additional hydrogen was produced. Concentrations of both these substances had to be considered in calculating G(02 ).

The accident considered consisted of a LOCA in which 5 percent of fuel cladding was oxidized by reaction with steam producing failure of all fuel rods and overneating of the core, but eithout initiation of fuel melting. This case represented maximum degradation of core allowed by 10 CFR 50.44(d)(1) and 10 CFR 50.46(b)(3). The analyses performed by Sandia (Reference 5), based on the experimental work on fuel rods from the H. B. Robinson plant, have indicated that for this type of accident 30 percent of total fuel iodine inventory was released. The released iodine consisted of the initial gap inventory and of the iodine diffused from the overheated fuel. Assuming that all the releasted iodine was dissolved in water and using plant parameters corresponding to a typical BWR with Mark I containment, the iodine concentration in water was determined to be 1.11 E-5 moles / liter and the partial pressure of hydrogen in the containment 0.12 ata. This partial pressure corresponds to an equilibrium concentration of 1.9 cc hydrogen /kg of water. Inserting this value of iodine concentration into the BNL sathematical model a relationship between G(0 2) and partial pressure of hydrogen in the containment was developed. From this relationship it was determined that for a,non-boiling case,-when partial-pressure of hydrogen was 0.12 ata., G(0 3)=0.19. It also found that G(03 )

would not reach zero value until partial pressure of hydrogen in the containment reaches 1 ata. For boiling case, when hydrogen is stripped from the solution, G(0 )3 would be slightly higher, somewhere between 0.19 and 0.22.

tw.

a TR-081 Rev. 0 Page 46 These values differed considerably from those in the NEDO-22155 report. The main differtoce was probably due to the GE results being applicable to pure water or to water containing only minimal amount of impurities. Including the  ;

effect of iocine, which wovid be released during certain types of LOCA, could '

drastically change the resolts, CCNCLU5!CNS AND RECCMMFNDATIONS

1. The NE00-22155 report underpredicts generation of radiolytic hydrogen for both boiling and non-boiling cases. This is due to the use of too low values for G(03 ). G(03 )=0.1 for boiling case was based on the measurements made in an environment of zere or low todine concentrations. G(02 )=0 for non-boiling case was derived from the data calculated by the codes which did not consider effects of dissolved iodine. The results were also in disagreement with the experimental data from ORNL.

2. Since G(02 ) is a function of hydrogen and iodine concentrations in the containment water, it may vary during an accident and is specific for each individual plant.

3. The maximum values of G(0 2 ), calculated with the NRC radiolysis model for LOCA (5% metal water reaction and 30% iodine release) in a BWR with Mark I containment, are G(02 )=0.19 for non-boiling and between 0.19 and 0.22 for boiling cases. They are considerably higher than the values presented in the General Electric's NE00-22155 report.

4 The value of G(02 )= 25 in Regulatory Guide 1.7 is overly conservative.

However, it is not very much different from the maximum values calculated for a LOCA using the BNL model. It is recommended, therefore that until a better understanding of post accident radiolytic decomposition of water is developed, this value should be used for predicting generation rates of radiolytic cxygen in the containment.

REFERENCES

1. J. C. Conine, D. J. Krommenhoek and D. Emanual Logan, KAPL Evaluation of i

Radiolysis Associated with Three Mile Island Unit 2 Incident, dated May 1979.

2. 5. Gordon, K. H. Schmidt and J. R. Honekamp, An Analysis of the Hydrogen Bubble Concerns in the Three Mile Island Unit 2 Reactor Vessel, Argonne I National Laboratory.

l 3. NRC Memo and K. I. Parczewski to Victor Benaroya, dated June 23, 1982.

A TR-281 a Fev. 3 4 Page 47

4. H. E. Zittel, Design Considerations of Reactor Considerations Spray Systems - Boiling water Reactor Accident Studies, ORNL-TM-2412, Part VIII, Octoeer 1970.

S. NuREGrCR-2367, updated Best-Estimate LOCA Radiation Signature, date'd August 1991.

Principal Contributor: (. P arc zews ki Dated: July 6, 1989

_ -____-__ - - _ _ _ _ _