ML20203L954

Iodine & Cesium Releases Due to ECCS Leakage

ML20203L954
Person / Time
Site:	Byron, Braidwood, 05000000
Issue date:	12/31/1984
From:	FAUSKE & ASSOCIATES, INC., SARGENT & LUNDY, INC.
To:
Shared Package
ML20203L889	List: ML20203L887 ML20203L954
References
FAI-84-59, NUDOCS 8609020096
Download: ML20203L954 (19)

Text

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

f.

7

.t s

4:

J J

2 ATTACHMENT II l

IODINE AND CESIUM RELEASES DUE l TO ECCS LEAKAGE i-

j. Fauske and Associates,'Inc. t i Sargent & Lundy ,

J i

1.

1 4

i P

i

$ I f

~~

8609020096 860826 l

i kDR ADOCK 05000456 PDR ..

1 1

- , ,,.-,,v-wr,.-.vcmy*,en-w -v.,-,. y-'.-==ww-.,e,y,-+-,.e--,.,m.v..w-.,--w.._ .,,,-vvw yc.,mw-w,wv-r----+-r,----y--ww,,yw,mw.-

f -

4 '

, FAl/84-59 l

E

( -

D Fauske & Associates, Inc.

16WO70 West 83rd Street

,- Burr Ridge, Illinois 60521 (312) 323-8750 i

l IODINE AND CESlUM RELEASES DUE TO ECCS LEAKAGE Sargent & Lundy December,1984 i

t c .

1.0 INTRODUCTION

Licensing requirements for the Byron fiuclear Power Station required an assessment of the potential leakage of iodine and cesium into the auxfifary building environment for design basis accident conditions. The leakages to be considered are 1 gpm for a 30 day period and 50 gpm over a 30 minute interval as a result of potential leaks in charging pumps. RHR pumps, etc. The re-

~ 1 cases of iodine and cesium within the core are those assumed for licensing cases, i.e. half of the core inventory of each of these rcdicactive elaments.

For these analyses, the chemical state is considered to be iodine gas and

( elemental cesium as opposed te the dominant and 'less vol tile states of cesium iodide and cesium hydroxide.

Consequently, a considerable conservatisn is introduced into the analysis as a result of there assumed chemical states.

l The radioactive elements are assumed to be lost from the fuel matrix an deposited within the emergency cooling water, which includes the primary, system water as well as +he refueling water storage tank (RWST). Corsidering half of the iodine to be released from the core, about 0.06 kg/ moles of fodine (7.5 kg) would go in the solution with 1.8 x 106 kg (484.000 gallons) of wa ter.

The questions addressed will be the concentration of fodine and cesium j

within the water and its removal from the water by diffusion into the circu-l lating air within the cell. This is addressed for both leakage conditions.

I

• It is recognized that elemental cesium could not te in solution with water.

Therefore, evaluations are also carried out for the most likely chemical forms of cesium iodide and cesium hydroxide dissolved in the water.

t ,

i -

, , - . . , - ,,,,,--,,---..---m-. _

---

,..,n-.,. ._,w..---m,--,..,-,,. -,..,.._n,m-. ,.,.w _..,,,,--,,_..,,.,,---n---, - , - - - - . - , ,

2.0 BASIC CHARACTERISTICS FOR CESIUM AND IODINE In the initial analyses carried cut in this this report,, fodice and cesium will te assumed to behave as elemental species I and Cs. The vapor 2

pressures for these species are given by j Iodine: InP = - 6119/T + 24.81 (2.1)

, Cesium: InP = - 8513/T

• 20.35 (2.2) where tha pressure (P) is in Pascals and T is in degrees K. These will be used to . determine the partial pressure of these elements in an aqueous solu-tion. The masses of the two elements consfo'ered are 7.5 kg of 1 2 and 83 kg of Cs which is typical of an equilibrium core cycle for a 1000 MWe plant. This amount is assumed to be released into 1.8 x 106 kg of water during the opera-tion of ECCS functions forming an aqueous solution of the elemental species.

The effective partial pressure of the iodine or cesium in the aqueous solution can be estimated through Aaoult's law. The expression for the partial pres-sure (PP) of a dissolved element (1) in solution is given by N-ppg = P3 ,g(T)

(2.3)

! where P sat (T) is the saturation pressure of tho element at the solution temperature (T), Ng are the moles of iodine in the solution and N are the total moles. 7 .

At equilibrium, the partial pressure of the dilute species in the gas phase is equal to the effective partial pressure of the species in ' aqueous soluticn.

Generally, this has been characterized as a partitioning coefff-cient (H) dufined as the ratio of the mtterial concentration in the liquid (Cg ) divided by the concentration in the gas phase (C ).

g H = Ct/Cg (2.4) 9 n,--_-..

The concentration in the gas phase can be characterized as the mass of the dilute species divided by the total gas volune, and assuming' the species behaves as an idaal gas, this can be expressed as .

PP4 Mw 4 C a g g (2.5) where % is the molecular weignt of iS species, R is the universal gas constant and T is the absolute temperature. Similarly, the concentration in the liquid phase is the mass of the material in aqueous solution divided by the volume of the water as expressed by i

I Ng MGgo CL" NwMww (*

where N,, M, and o, are the number of moles. the molecular weight and density l .

of water respectively, Using the Raoult's law for the effective pressure of the material in solution, and assuming N, = N T

, the partitioning coefficient under equilibrium conditions can te expressed as This predicted behavior can be compared to the measurements in the CSE experiments [1] in which elemental and particulate iodine was injected into a steam atmosphere along with cesium, uranium and ruthenium. The materials were.

! accumulated in the sump water due to steam condensation, gravitational set-l tling and direct vapor condensation. After about 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />, the ratio of the concentration in the sump water to the concentration in the gas base remained essentially constant. This value can be compared with the prediction from Eq.

(2.7) to demonstrate the viability of using solution chemistry to evaluat'e the effective partial pressure of the various species in aqueous solution. Such a j

comparison is given in Table 2.1 and is seen to be in general agreement with the measured values, with the deviation between the experiments and predic-tions showing the material to be more tightly bound in thi sclution than predicted by the simple model. This is generally attributed to reactions in the water that make the iodine less volatile than the consideration of

.--. - - . . _ - . - - - . _ _ . . _ ~ . .,..._.___,,..,....____,._._.____._-.,..,______._.._.__m.- - , _ . - - - . . . _ - -

. . 1 Ta 31e 2.1 COMPARISON WITH CSE IODINE RESULTS T Time C C H H Test

K Hrs. ufi g ug 1 Experimental Predicted A-1 356 4 1.8 (-3)* 1.8 (2')

. 1 (5) 8.1 (4)

8 5 (-4) 1.2 (2) 2.4 (5) 8.1 (4) 12 3 (-4) 8 (1) 2,7 (5) 8.1 (4) 24 1 (-4) 4 (1) 4 (5) 8.1 (4)

A-2 358 4 7(-2) 1.3 (4) 1.9(5) ,

7.4(4)

, 8 4 (-2) 9(3) 2.2(5) 7.4(4)

! 12 3 (-2) 7 (3) 2.3 (5) 7.4(4) 24 1 (-2) 4 (3) 4 (5) 7.4 (4)

A-5 396 4 1.3(-1) 2 (4) '1.5 (5) 1.6 (4) l 8 7 (-2) 1.3 (4) 1.9(5)- 1.6(4) 12 6 (-2) 9(3) 1.5 (5) 1.6 (4) 24 4 (-2) 7 (3) 1.8 (5) 1.6 (4)

A-11 395 4 1.9(-1) 1.4 4) 7.4 4) 1.6 (4) 8 8 -2) 1.3 4) 1.6 5) 1.6 (4) 12 5 -2) 1.2 4 2.4 5) 1.6 (4) 24 3 -2) 1 ( )) 3.3 5) 1.6 (4)

• 1.8 x 1,0-3 .

I e

m

,,---,---,-,-v--,-w,

4 elemental iodine. As a result, the effective partial pressure of the species I  !

in an. aqueous solution would be from simple solution chemistry. It should ce remembered that in this experiment, elemental fodine was infected directly into the' gas base as opposed to the postulated accident case in which material removed from the fuel matrix would have the dominant chemical forms of cesium '

iodide and cesium hydroxide. Therefore, the comparison with the CSE experi-ments should be viewe.d as a qualification of the solution chemistry approach 1

and its conservatisms, but not an indication of the dominant chemical states.

A similar calculation can be carried out assuming that cesium could exist

~

in solution, which is conservitive, but not physically possible. For the CSE tests the gas phase did not contain measurable cesium vapor, but cesium

hydrcxide in particulate form.

With this approach to the effective vapor pressure, the driving force for diffusion of vapors from the ECCS leakage can be estimated for the cor.ditions of interest. Since the spills onto a cubicle ficor would be removed to holding tanks via ficer drains, the key element of the analyses is the rate l

dependent process of diffusion from the liquid into the gas phase over the .

time interval of interest, i.e. 30 minutes .for a large spillage and 30 days for the 1 gpm leak rete.

I 1

e

, i s

J v --~, - - m- .---w- , na, --, ------~n..,,--. ----,----e ---,-----,-----..%- - - ~ . - , - - - - , .n----- - , .-------. ._.---_,-.e-,-------

~

d 3.0 RATE DEPENDENT PROCESSES s

Th'e diffusion from the spilled liquid can be estimated from t 0 FP, N /A = R 6 g

(3'I) where N g is the rate of moles diffused for the iS species, A is the surface

~

area for diffusion. O is the diffusivity in the gas phase, R is the universal gas constant, T is the absol'ute temperature of the liquid, PP g is the partial i

pressure of the species in aqueous solution and 4 is the diffusion boundary layer. In this calculation, the diffusion rats dependent process in the

{ liquid phase is ignored as is the partial pressure of the species in the surrounding gas phase. Both of thest make the analysis socewhat conservative, i.e. the diffusion rate will be overestimated. Table 3.1 gives the asstap-tions of material qnntities for core inventory, water inventory and the concentration of the iodine should complete ionization occur. Table 3.2 lists the assumptions in the FSAR analysis (90% retention by the liquid and 90%

filtration by the auxiliary building ventilation system) to estimate the

{ quantities of iodine mass lost to the environment. This value of 0.15 g of l fodine can be compared to the calculations resulting from evaluating the rate dependent processes.

l Figure 3.1 shows an assumed configuration for a continuous leakage of 1 l gpm for 30 days ~in which the stream would pour onto the floor and into the floor drains. As a result of the leakage, the stream can be viewed as a j continually resupplied column of liquid such that the concentration of the dissolved species is constant over the interval of interest. For these calculations., a diffusion boundary layer of 2 mm is assumed in the gas phase, and the material pouring onto the floor is assumed to have an average velocity of 1 m/see to account for run off on equipment and running onto the floor.

With this velocity, the effective diameter of the streem is about 1 on and heat transfer coefficients show that the stream temperature would decrease only slightly during its resonance time in the equipment cell. Consequently, p

a temperature of 373'K is assumed for the water during its residence interval.

l Table 3.3 lists the saturation pressure of iodine at this temperature, the

e Table 3.1 ASSUMPTIONS Core inventory 1 - 15 kg Cs -166 kg Released from Fuel 1 - 7.5 kg Cs - 83 kg 64,740 f t3 == 484,000 gal.

~

Water inventory -

1.8 x 10 6 kg = 10 5 kg moles Moles of I = 0.058 kg moles

~7 Concentration = 0;055 = 5.9 x 10 Spillage ~ 1000 gal. = 3636 kg = 202 kg moles

= 1.2 x 10

~4 kg moles l ~

e l

J l

1.________._ ______ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ . _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ . _

1 l Table 3.2 FS AR AN ALYSIS Westinghouse FSAR Analysis = 1.2 x 10

-5 kg moles i

released to the ,

auxiliary building atmosphere

= 1.2 x 10 -0 kg moles l released to the l

l environment

= 1.5 x 10 -4 kg

= 0.15 g p..

< J .

--ep.,--, ,- e- -,-w _mp,ma,,-__-- , n.,m.,.,m-e_-, ,w-,, -<ww_ - - _ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

f .

CONTINUOUS LEAKAGE l

1

= >

l V

i '

(1 gpm)

(30 days) l l

Fig. 3.1 Assumed configuration for ,

~

continuous leakage. 89 E"

< J .

,--,m-- -n-- ,-,--,4~,.,,--..n,--,,- -- -,-,-w.,a ----------m,,w -

l Table 3.3 '

CONTINUOUS LEAK AGE (1 com for 30 days) i Assumed Surface Area = 0.063 m Diameter of Water Stream ~ 0.01 m Interval = 2.6 x 106 seconds T =~ 373 K Pg(T) = 4470 Pa lodine PPg (T) = 2.6 x 10 -3 p, N/A = 4.2 x 10 -12 kg moles /m /sec

, 6, = 3.4 x 10 -11 kg/sec Am, = 8.7 x 10 -5 kg = 0.087 g Cesium lodide

-13 Peg (T) = 2.1 x 10 p, PPeg(T) = 1.2 x 10 -18 Pa N/A = 2 x 10 -28 kg moles /m2 /sec

= 3.3 x 10

-27 kg/sec Sci .

-21 Amg= 8.5 x 10 kg e

E

< J l

I l

l

I partial pressure of the material in solution, the diffusion rate per unit area, the rate of iodine lost to the atmosphere and the total lost over the entire -30. day interval. As shown, this value is comparable to..but less than

[ the value used for the FSAR analysis when the building ventilation system is credited for removing 90% of the airborne iodine. This calculation demon-strates that the solution chemistry is very effective in retaining the fission products and would not release substantial quantities to the cubicle environ-ment, even when elemental iodine is assumed as the chemical state. Table 3.3 also illustrates the influence of the chemical state by demonstrating the amount of mass lost for cesium iodide held in solution, which essentially shows only negligible quantities to be released. Obviously, the actual release would be larger than that calculated assuming cesium fodide as a result of secondary reactions in the ' water, but the release value would be, less than that calculated assuming elemental fodine to be in solution.

Release calculations for cesium and cesium hydroxide are listed in Table 3.4,

, and these are well within the tolerable levels.

Figure 3.2 illustrates the configuration for a large spillage of 50 gpm over a 30 minute interval. The assumpticn is that the spill rate is suffi-cient to accumulate a layer of water on the cubicle floor before it runs off into the floor drain. As a result, the surface area for diffusion is equal to the cubicle floor area and is taken to be 30 m2 in this analysis. While the surface area is larger, the time available for diffusion equals the spill time of 1800 seconds and Table 3.5 lists the assumptions and the results for assuming the chemical state is fodine and cesium iodide. The net result of assuming elemental iodine and solution is a release quantity which is approxi-mately 20% of that calculated in the FSAR and the assumption of cesium iodide results in an extremely small amount of material released. Table 3.6 provides calculations ,for the assumptions of elemental cesium and cesium hydroxide l under the same conditions, and as shown, reveal that cesium would be tightly bound in the water regardless of the assumed chemical state. It should again be noted that elemental cesium could not be in solution with water and cesium hydroxide is the dominant chemical state.

These analyses demonstrate that the assumed spill rates do not result in substantial release of iodine or cesium to the cubicle envirorment even if the l

l

, Table 3.4 l

CONTINUOUS LEAXAGE T = 373 K Cesium p (T) . o,og4 p, ,

PP (T) = 5.3 x 10'7 Pa N/A = 8.5 x 10-16 kg moles /m2 73,,

$c = 7.1 x 10-15 kg/sec am e

= 1.9 x 10-8 kg l

Cesium Hydroxide l

I P CSOH(T) = 9.1 x 10~9 PP CSOH(T) = 5.7 x 10*I4 Pa N/A = 9.2 x 10-23 gg ,,j,,f,2/sec ACSOH = 7.7 x 10* N kg/sec kg AmCSOH = 2 x 10 l

I l

l

(

• l LARGE SPILLAGE l

I w

i

)

V A A A- A ll V

. 50gpm 30 min.

Fig. 3.2 Assumed configuration for a large W spill on a cubicle floor, g4 .

Il .

( )

l.

1 Table 3.5 .

l LARGE SP!LLAGE j (50 com for 30 minutes) l .

l RHR Pump Cubicle 20' long x 15' wide 300 ft2= 30 m 2 Area =

Mechanism - Diffusion from the water pool with an 2

area of 30 m and an interval of 30 min.

1 i . T = 373 K lodine Pg(T) = 4470 Pa

! PPg (T) = 2.6 x 10 -3 p, N/A = 4.2 x 10 -12 kg moles /m2 j,,,

mg

-8

= 1.6 x 10 kg/sec

-5 Amg = 2.9 x 10 kg = 0.029 g Cesium lodide -13 p, Peg (T) = 2.1. x 10 ,

i -18 Pa PPeg(T) = 1.2 x 10 N/A = 2 x 10 -28 kg moles /m 2 7,,,

l S eg = 1.5 x 10 -24 kg/sec W Ame g = 2.7 x 10 # kg [

( J

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

Table 3.6 -

1 LARGE SPILLAGE

. T = 373 K

- Cesium Pe (T) = 0.084 Pa ,

PPc (T) = 5.3 x 10 ~I Pa N/A = 8.5 x 10 -16 kg moles /m 2 7,,,

-12 kg/sec se = 3.4 x 10

-8 f Ame = 6 x 10 kg

~

j Cesium Hydroxide PCsOH(T) = 9.1 x 10 ' Pa

~I4 PPCsOH(T) = 5.7 x 10 Pa

! N/A = 9.2 x 10 -23 kg moles /m 2 7,,,

mCsOH = 4 x 10-18 kg/sec

-16 kg A*CsOH = 7 x 10 c v j

I J

• elemental form is assumed to be in aqueous solution. Several conservatisns are inherent in the analyses as carried out and these are delineated in Tab'e 3.7. The first is that an equilibrium core cycle was assumed a,nd this couad overstata.the amounts of fission product material available in the early stages of the Byron plant operation by a factor of 2 to 10. In addition, it is assumed that 50% of the iodine and cesium fission product material are re-leased from the fuel matrix, which for a design basis accident could overstate l

the tission products in aqueous solution by a factor of 10 to 100. The ,

comparisons with the CSE experiments demonstrate that the partition coefff-cient is larger than calculated by simple aqueous solution chemistry, which

! could potentially decrease the rates of materials lost to the environment by a factor of 2 and perhaps as much as 10. Lastly, the dominant chemical states, which is not an independent change from the partitioning coefficient listed in Item 3 decrease the release by orders of magnitude. Considering low and of the significance of these conservatisms and neglecting the dominant chemical l

state, the analyses are conservative by at'least a factor of 60 if more realistic assessment's were applied to the actual core cycle, the materf at-I released from the fuel matrix and use of an experimentally determined parti-i . tioning coefficient from the CSE tests. However, the analyser already show that the efficient retention of fission products in aqueous solution are equal to the releases considered in the FSAR, hence further refinement of the calculations would not appear to be warranted.

e 4

S e

l

l [ , . \

Table 3.7 i

CONSERVATISMS IN THE ANALYSES ,

l .

Type Significance

1. Equilibrium Core Cycle 3 - 10

2. 50% i and Cs Released From 10 - 100 the Fuel l

3. Neglect Partitioning Coefficient 2 - 10 for lodine ,

4. Chemical State Cal and Cs0H Several Orders of Magnitude

~

j .

l i

E l t >  ;

1

~ .

4.0 REFERENCE

1. R. K. Hilliard and L., F. Coleman, " Natural Transport Effects on Ffsston Product Behavior in the Containment Systems Experiment," BNWL-1457, December. 1970.

4 i

e f

4

(

0 O

e o

t -

/

-+ . .- . --. . , -....w -- --

,--- - + - , ,--- . - . - - - - , , . .. +,,- - w . ----,w- - - - , --. ----------,y-, , - - - - - - - - -%-,.---, ,---- ~ww