ML20206T799

Iodine Revolatizitation in Grand Gulf Loca

ML20206T799
Person / Time
Site:	Grand Gulf
Issue date:	01/31/1999
From:	Beahm E, Carl Weber OAK RIDGE NATIONAL LABORATORY
To:
Shared Package
ML20206T717	List: ML20206T799
References
RTR-NUREG-1465 ORNL-M-6544, NUDOCS 9902120055
Download: ML20206T799 (19)

Text

. 7

)

ORNL/M-6544 Computational Physics and Engineering Division I

l 1

IODINE REVOLATIZITATION IN A GRAND GULF LOCA C. F. Weber and E. C. Beahm Date Published: January 1999 Prepared by the OAK RIDGE NATIONAL LABORATORY P. O. Box 2008 Oak Ridge, Tennessee 37831-6370 managed by LOCKHEED MARTIN ENERGY RESEARCH CORP.

for the U.S. DEPARTMENT OF ENERGY under contract DE-AC05-960R22464 9902120055 990210 PDR NUREG 1465 C PDR

)

t CONTENTS t

l l Pagt L

LI ST O F TAB L ES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v 1

l' LIST OF FIGURES ........................................................vii 1

1. INTRODUCTION ..................................................... I

2. DATA AND AS S UMPTIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3  !

4 1

3. CHEMISTRY MODELS ........... .................................... 7 I i 3.1 DETERMINATION OF pH ......................................... 7 3.2 IODINE CHEMISTRY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

4. RESULTS AND CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 i

4

5. RE FERENCES . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 I i

l i

i l

1 l

i til

l t

LIST OF FIGURES i

/

1 1

Figure Page '

l. Radiation Doses to Water Pools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 3

2. Radiolytic Acid Generation ...... ............. ......... .. ........... 13

3. Water Pool pH . . . . . .......... ..... . ...... . . ...... ....... 14 l l

l

4. Iodine in Containment Gas Space . . . . . ...... .................. .. 15

5. Iodine Released to the Environment . ... .......... . .................... 16 l

l l

V f

]

LIST OF TABLES 1

Table Page

1. Control volume data . . . . . . . . . . . . . . . . . . ........... ............... 4

2. Inter-compartmental flows .............................................. 5

3. Nuclide group inventories and dose rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1

0 I

l l

t Vil

4 1

1. INTRODUCTION  ;

l Evaluation of reactor severe accident sequences has generally assumed that the dominant form ofiodine is CsI.' Emitted from the reactor coolant system as an aerosol, the iodide easily deposits on surfaces or enters water. Deposited parides may be washed into sumps or pools if '

water films from sprays or condensation are preseut c.<. the containment surfaces. Otherwise, the deposited species are fairly stable, and unlil.ely to move.

In water, Csl readily dissolves into the constituent ions Cs+ and 1, both of which are quite stable. However, under certain conditions I could react with various other aqueous species and form non-trivial amounts of1.2 While some 1 2will stay dissolved, it evaporates from water much more readily than other species. Once in the gas space,it may be carried with leakage or venting flow into the environment. Thus, the aqueous conversion of1 to 12 Presents a significant additional threat of environmental release.

It has been shown 2Jthat the conversion ofiodide to molecular iodine takes place only in acidic solutions. At 25 C, the reaction is quite slow for pH above 5, and quite rapid for pH under 3. At 90 C, the reaction procedes rapidly for pH < 5. Thus, the revolatilization ofiodine is dependent primarily on pH and, to a lesser cxtent, temperature.

This study is an attempt to quantify iodine revolatilization from water pools using mechanistic models. A Loss of Coolant Accident (LOCA) at the Grand Gulf Nuclear Station has l been chosen for the sample calculation. This sequence has been recently modeled using the  ;

MELCOR code / forming a basis for this analysis. As is the case in all Boiling Water Reactor  !

(BWR) plants, the pressure suppression pool at Grand Gulf has no pH control; hence, there exists i the real possibility oflowered pH and iodine revolatilization.

Under Nuclear Regulatory Commission (NRC) guidance, models ofiodine behavior under various severe accident conditions have been developed at Oak Ridge National Laboratory (ORNL). These models have come to be known collectively as the ' TRENDS' code. NRC support of model development ended in 1992, t efore any formal documentation or code development was undertaken. However, at the same time a need arose to actually apply most of the iodine chemistry and interaction models to postulated severe accident sequences at the High Flux Isotope Reactor (HFIR) in Oak Ridge. As a part of this work, substantial documentation was accomplished.5 In addition, the code HFIR-TRENDS was developed to actually perform the calculations. This code is not easily exportable or transportable to other plants or sequences, inasmuch as many statements are tailored to the particular plant (HFIR) for which they were written. Later, certain parts of HFIR-TRENDS were appropriated and modified so as to simule.-

three Westinghouse AP-600 design basis accidents.63 Again, the code was specifically tailored to the plant and accident sequence.

The TRENDS models do not include primary system, aerosol, or thermal hydraulic analyses. In the past, basic plant analyses were done using the MELCOR code for the HFIR analysis and MAAP code for the AP-600 reactor. Transient variables such as temperatures, pressure and fission product inventories were then input into the TRENDS analysis. Likewise, the TRENDS analysis for Grand Gulfis based on the recent MELCOR results, which will be described in more detail in the next section.

I

2. DATA AND ASSUMPTIONS As mentioned, the TRENDS calculation does not perform the basic accident analysis, but rather imports such information as flow rates and temperatures. In the present study, all such information was obtained from Ref. 4. The nodalization for the TRENDS analysis is exactly the same as for the underlying accident analysis. The control volumes are listed in Table 1, together with their total volumes, and gas-lic, 'd interfacial areas. [ Interfacial flow areas were obtained from the volume look-up tables by assuming a completely open top surface; if the volume Vis a linear function of elevation h, the interfacial area is A=(V 2 - iV )/(h:- ih ).] Almo:t all flow paths used in the TRENDS analysis also correspond to analagous flows in the MELCOR calculation, -

listed in Table 2. 1 The TRENDS analysis requires a calculction only every two to three minutes during the i transient. If MELCOR time steps occur at roughly this interval or greater, they are followed exactly. If MELCOl. times are more frequent, then data from multiple times are combined so as to produce a single r nge value for a TRENDS time step. At each time step, the following transient information from MELCOR results are imported into TRENDS:

1) Temperatures (liquid and gas) for each control volume

2) Flow rates (liquid and/or gas) for flow paths between control volumes

3) Liquid volumes (where not zero) l

4) Deposition rate of Csl into water pools l

5) Pressute in various regions of the reactor containment I i

l

6) Dose rates to water pools and airspaces.

The dose rates themselves are calculated as follows. MELCOR results provide inventories (kg) of each fission product group in each control volume at each time step. Molar dose rates for each group (MeV/s mol) were obtained during the previous study6 using ORIGEN results for individual nuclide inventories in the AP-600 reactor, combined with ICRP decay data. These are shown in Table 3 for both p and y radiation. (While ORIGEN calculations reflect assumptions of burnup and fuel composition for the AP-600 sequence, it is expected that the results do not differ substantially from what would be expected in Grand Gulf. In any event, such distribution data are not available for Grand Gulf, and wouS require additional time and expense to obtain them.) Multiplication of the molar dose rate by group molecular weight and group mass produces the dose rate contribution for the fission product group.

The dose rate for each control volume is then obtained by summing the dose rates for each fission product group within the volume. The resulting dose rate in water pools is shown in Fig.1. Similar results are also available for the drywell airspace and containment airspace.

3

The refueling pool was not considered in the present analysis for several reasons. All distinct flows are outward, however water enters through the " rain" of containment sprays. The pool eventually fills up and overflows into the pressure suppression pool. Within a few hours almost all fission products are in the latter, even those that migrated through the refueling pool first.

Finally, the source terms for this calculation [which determine the Csl deposition in 4) above] are based on NUREG-1465,8 as described in ref. 4. Total amounts added are stated in Table 3, riong with the energy deposition rates for each group.

l Table 1. Control volume data  ;

Volumes (m)) Interface area 2

Index Description (m )

)

I dry-head-seal 133.1 0. O.

2 pedestal-cavity 259.5 .15 32.7 3 weir-wall-ann. 462.9 353.8 62.5 4 drywell-ann-4 1777. O. O.

5 shield-wall-aan. 380.1 0. O.

6 drywell-ann-3 1319. O. O.

7 drywell-ann-2 2867. O. O.

8 drywell-ann-1 1165. O. 230.4 9 drywell-head 161.5 0. 38.1 10 suppression-pool 7989. 3509. 619.1 11 wetwell-ann-1 3712. O. O.

12 upper-pool-ann-1 1625. O. O.  ;

l 13 refuelpool 2154. 2067. 285.3 14 upper-ctmt-regl 3535. O. O.

15 upper-ctmt-reg 2 3535. O. O.

16 dome-h; aisphere 10560. O. O. l

17 ww-ann-lower 2392 0. O.

18 ww-ann-upper 2392. O. O.

19 upper-pool-ann-2 1707. O. O.

20 pump-sampling 354.7 0. O.

21 main-steam-room 475.7 0. O.

22 rwcu-hx-room 297.3 0. O.

23 valve-nest-room 90.8 0. O.

24 rwcu-filter-dem 129.5 0. O.

25 rwcu-pump-room 21.8 0. O.

26 rwcu-tank-room 97.6 0. O.

27 equip-hatch 1582. O. O.

28 enclosure 16990. O. O.

29 SGTS 30 enviro:unent 31 reactor vessel 4

Table 2. Inter-compartmental flows Index Source Sink MELCOR-ID 1 31 2 19+20+21 2 31 8 22+23 3 31 10 100+102 4 8 7 201 5 7 6 202 6 6 4 203 7 4 1 204 8 5 1 205 9 9 1 206 10 2 5 207 11 8 2 208+209 12 4 19 210 13 3 10 211+212+213 14 7 3 214 15 7 10 215 16 9 1 216 17 10 11 301 18 11 12 302 19 19 17 303 20 17 18 304 21 17 14 305 22 18 15 306 23 18 16 307 24 15 16 308 25 14 15 309 26 13 14 310 27 12 19 311 28 28 30 312 29 19 20 313 30 22 21 314 31 19 22 315 32 13 10 316 33 12 23 317 34 19 24 318 35 12 25 319 36 25 26 320 37 10 27 321 38 27 11 322 39 27 12 323 40 27 19 324 41 27 17 325 42 18 28 399 43 16 28 400 44 28 30 401 45 28 29 402 5

Table 3. Nuclide group inventories and dose rates l

l Specific dose (MeV/s) / mol

• 10 "

Total l Nuclide Release gamma beta Molecular Group (kg) 2 hr 10 hr 2 hr 10 hr Weight I Xe 463.7 8.678 2.964 6.098 3.718 131.3 2 Cs 61.55 13.99 5.261 14.33 2.25 149.913 3 Ba 4.15 27.31 11.5 32.76 18.99 137.34 4I(as1) 2 .30 1234 687.8 365 217.2 253.8 5 Te 2.04 98.01 37.34 46.86 18.65 143.6 6 Ru 0.77 10.19 9.167 9.167 7.324 101.07 7Mo 0.88 2.1 1.876 5.36 4.904 95.94 8 Ce 0.30 18.53 16.69 27.9 25.16 140.12 9 La 0.11 35.01 26.23 35.01 23.17 138.91 10 U - 0.1113 0.008796 0.0284 0.01208 270.03 16 Csl 12.85 1248 693.1 379.33 219.45 259.8 Aerosol 17 I .01 1234 687.8 365 217.2 253.8 organic vapor l

l l

6 m

3. CIIEMISTRY MODELS The TRENDS analysis models certain chemical reactions not considered in major accident I analysis codes such as MELCOR and MAAP. This additional analysis centers on two areas of importance: determination of pH in water pools, and speciation ofiodine.

3.1 DETERMINATION OF pH The calculation of pH is performed in each control volume at each step, immediately prior to the iodine speciation calculation. It assumes that the system is constantly in chemical equilibrium with respect to species that significantly affect pH. Internal coding ensures that time steps are small, so that changes in chemical inventories of each control volume are also small.

The actual equilibrium computation is performed using the principal subroutines of the l SOLGASMIX code, as described in refs. 5 and 6. With no borates or phosphates, the only significant additives are fission products and acids generated radiolytically in containment airspace and deposited into water.

The rate of hcl production is based on energy deposition rates (both p and y) due to fission products that are airborne or deposited on surfaces. Acid production rates are calculated from the procedure described in ref. 3, Appendix B. This empirical model is based on dose rates (described in Sect. 2) and amount of cable insulation present (estimated as 9835 lb in the drywell and 1.764 x 10' lb in the containment). Since most flow in containment tends to be downward toward the wetwell water, all acids produced in containment were assumed to deposit in the wetwell water immediately. Acids produced in the drywell were routed to the sump during the first 20 minutes and afterward to the wetwell (consistent with water and gas flows).

The irradiation of air-water systems also is known to produce nitric acid. However, the quantity expected is dwarfed by the hcl production and hence, was not considered 1n the present study.

3.2 IODINE CHEMISTRY  !

l While all iodine enters water pools as iodide (I-), it may change to more volatile forms during the course of the accident. The iodine species distribution is determined by solving rate equations described below.

I 3.2.1 Hydrolysis In aqueous solution, iodine can undergo hydrolysis which is described well by the reaction set" 7

9 12 + H2 O - I + HOI + H * (2a) 2 HOI - I + HIO2+H' (2b)

HOI + HIO 2

- I + HIO + H ' .

3 (2c)

The endpoint oxidation states I and IO-(or HIO )3 are both highly soluble, and therefore highly desirable. The species HOI and HIO2are reaction intermediates which may be volatile, but which are regarded as having short lifetimes. However, molecular iodine 12 is stable and sparingly soLble. It can be seen in the above reactions that low pH (i.e., large H* concentration) will result in more 1,2 and less I and IOi . Each forward and reverse reaction gives rise to a single rate equation, as described in ref. 5.

3.2.2 Radiolysis 1

Under radiation, the presence of free radicals induces additional aqueous reactions.

While not as well understood, the overall result can be described empirically by the equilibium 5 (H *7(I-7 = a + b (H *) , (3)

(1) 2 where a and b are known coefficients depending only un temperature. The actual rate equations used are based on the catalytic decomposition of hydrogen peroxide, one of the stable byproducts ofirradiating water. The peroxide inventory is estimated empirically based on cumulative dose, pH, dissolved oxygen content, and temperature. The forward rate coefficients are then based on the data of Liebhafsky, and the reverse coefficients chosen to move steadily toward equilibrium (3).

3.2.3 Gas / Liquid Partitioning The same equation is used to describe evaporation of volatile iodine species from water (I 2or CH I)3 or their dissolution from gas. The model is based on the equilibnum partition coefficients (inverse of Henry's Law constant) for each species, and rate equations obtained from natural convection correlations.5 3.2.4 Organic Iodides Organic iodides are described empirically as the formation of CH3 1 from 2I in the gas phase. The model used is described fully in ref. 5. As is 1,2 CH 31 is sparingly soluble in water, an effect which is negligible in the present sequence.

8 s-

4

4. RESULTS AND CONCLUSIONS The TRENDS models are applied at each time step to each control volume. Significant amounts of water occur only in the wetwell and drywell sump (the refueling pool is not a factor, as discussed earlier). In Fig. 2, we show the radiolytic acid production feeding into each of these pools. Since the water is initially neutral and no chemical additives are present, the acid additions are the major factors affecting pH. In Fig. 3, we see the downward trend of pH resulting from these acid additions.

The conversion ofiodide (I-) to molecular iodine (1 2) is most noticeable in the wetwell, since this is the repository of most iodide and hcl. Gradually, during the transient small I I

amounts of more volatile iodine are formed. While iodide remains the dominant form, noticeable amounts of1 2and intermediate species are created.

Once produced in water, some I2 is free to evaporate into airspace. Fig. 4 indicates the increase in all airborne iodine throughout the transient. This is compared to the MELCOR result for Cs! aerosol, which decreases dramatically due to containment sprays. The 12 in the airspace can be vented to the enclosure building or the environment. In the present accident sequence, the only path to the environment was through the SGTS, which was assumed to operate as in MELCOR. However, both are dwarfed by the MELCOR gaseous release during the first 12 h because MELCOR does not model spray washout of gaseous iodine. Steadily increasing throughout the transient, the revolatilization release is eventually more than an order-or-magnitude higher than the MELCOR aerosol release. Also,99% ofiodine flowing directly through the SGTS was retained in filters. The remaining 1% was released to the environment.

In addition, a small flow bypassing the SGTS filters vented directly into the environment. The total released from these two paths is shown in Fig. 5.

9

5. REFERENCES

1. E. C. Beahm et al., lodine Chemical Forms in L WR Severe Accidents, NUREGICR-5732 l (ORNLTTM-11861), Martin Marietta Energy Systems, Inc., Oak Ridge National i Laboratory (1992).

I

2. C.-C. Lin, " Chemical Effects of Gamma Radiation on Iodine in Aqueous Solution,"

l J. Inorg. Nucl. Chem. 42,1101-7 (1980).

3. E. C. Beahm et al., Iodine Evolution andpH Control, NUREG/CR-5950 (ORNLITM-12242), Martin Marietta Energy Systems, Inc., Oak Ridge National l

Laboratory (1992).

4. J. J. Carbajo, MELCOR DBA LOCA Calculations, (Letter Report),

ORNL/NRC/LTR-97/21, Lockheed Martin Energy Research Corp., Oak Ridge National Laboratory, January 1999.

5. C. F. Weber et al., Models oflodine Behavior in Reactor Containments, ORNLITM-12202, Martin Marietta Energy Systems, Inc., Oak Ridge National Laboratory, January 1992.

6. E. C. Beahm, C. F. Weber, and T. A. Dillow, Technical Assistance in Review ofSource Term RelatedIssues ofAdvanced Reactors, NUREGICR-6408 (ORNLTTM-13144),

Lockheed Martm Energy Research Corp., Oak Ridge National Laboratory, October 1998.

7. C. F. Weber and E. C. Beahm, Iodine Volatility andpH Cbntrol in the AP-600 Reactor, NUREG/CR-6599 (ORNL/TM-13555), Lockheed Martin Energy Research Corp., Oak Ridge National Laboratory, October 1998.

8. L. Soffer et al., Accident Source Termsfor Light-Water Nuclear Power Plants, NUREG-1465, February 1995.

9. C. F. Weber et al., " Optimal Determination of Rate Coefficients in Multiple-Reaction Systems," Computers and Chem., 16(4),325-33 (1992).

l 10. H. A. Liebhafsky, "The Catalytic Decomposition of Hydrogen Peroxide by the l Iodine-Iodide Couple at 25 C," J. Amer. Chem. Soc., 54,1792-1806 (1932).

l 11

2 1

. 10 l

s o . 8 o

P l l

e r w t

e y _

a D r

W t

o )

s h

(

e s

o . 6 e D i m

n T i

o t

i a

d a

l l

R e t

w 1

. e . 4 e W r

u g

i F

. 2

- - - - 0 0 8 6 -

4 2 0 1 0 0 0 0 0

^. S X *$$$g:a $OO U

2 1

. 10 ll e

w y

l r l

e D t

w . 8 e

n W i

o t

a r

e n

e G )

d h

(

i c . 6 e A m c i i

t y

T l

i o

d a

R 2

e r , 4 u

g i

F

, 2

- - - - - 0 0 5 0 5 0 5 0 3 2 2 1 1 f }a D. gII$5EaO =

C _

\

2 1 _

. 10 l

l e

t w

e _

W _.

l l . 8 e

w y

r D

H p _

) _

l o h _

o ( _

P . 6 e -

r m e i t

a T W _

3 _

e r

ug i

F

. 4

. 2 0

7 6 5 4 3 co

2 1

' 0 1

s u

o e

s _

a _

G S

D _

- N '

8 e E _

c R _

a T p -

S -

s a

G )

r _

t _

n h

(

_ e m

6 e n

m i

a i

_ t T n

o

_ C

_ i n _

e i

n _

d 4 _

I o

. l 4 o s

_ e r

u m g e A

_ i F R _

O _

C '

2 R

M ' -

- - - - - 0 2 1 8 6 4 2 0 1

0 0 0 0 0 0 0 0 G 0 D2E$ 88 t c$ 5n.

2 1 -

s n

o -

it a

Sil z , 0 -

i Dt a R 1 Nl o Ol o Ev Cs Re L o Er e TR MA t , 8 n

e m

n o

r

. i v

n E }

e h

(

ht , 6 e t

o i m

_ d e

T _

s a

e l

e

_ R

_ e

_ i n ,

d , 4 o

I Rs 5

. Ouo e Ce L s r Ea u

_ i g MG F

, 2

- - - - 0 _

" 5-

• 4 4

• _

0 0 0 0 0 0 1 1 1 1 1 1 HEE 2$ o E@0 5a. '

E .

_