ML17054C230
ML17054C230 | |
Person / Time | |
---|---|
Site: | Wolf Creek ![]() |
Issue date: | 01/17/2017 |
From: | Wolf Creek |
To: | Office of Nuclear Reactor Regulation |
Shared Package | |
ML17054C103 | List: |
References | |
ET 17-0001 | |
Download: ML17054C230 (41) | |
Text
APPENDIX A TECHNICAL REPORT A-3788 8-12-86 FISSION PRODUCT REflJVAL EFFECTIVENESS CF CHEMICAL ADDITIVES IN PWR CONTAINMENT SPRAYS R. £. Davis, H. P. Nourbakhsh, and M. Khat1b-Rahbar Accident Analysis Group Department of Nuclear Energy Brookhaven National Laboratory Upton, NY 11973 August 1986 Prep a red for
- u. S. Nuclear Regulatory ColTITliss1on Washington, DC 20555 Under Contract No. DE-ACU2-76CH00016 FIN A-378ti
ABSTRACT The presence of gaseous iodine 1n severe accident situations 1s based upon 1 regulatory source term prescription whose basis predattS the accident et Three Mile Jsland-Un1t 2 and the source tenn research that TMI-2 stimulated. This report reviews the current best-estimate of source term characteristics, and the experimental bases that establish the effectiveness of spray additives. Based on this review, several current operating practices, vis-a-vis the addition of additive(s), may warrant regulatory reevaluation.
A-1
ACKNOWLEDGEMENTS The authors would like to thank Dr. J. Read (NRC) for the guidance he provided during the course of this effort. In addition. the authors appre-ciate the skillful preparation of the manuscript by Ms. Cheryl s. Conrad and Ms. Theresa Slteleney.
A-2
CONTENTS I. JNTRODUCTJON........................................................ 1
- 2. PAST AND CURRENT SOURCE TERM CHARACTERISTICS........................ 3-
- 4. SPRAY MODELS........................................................ 13 4.1 Analytical Procedure ******************************************* 13 4.2 Reevaluation of Existing Data ********************************** 15
- 5. DISCUSSION AND CONCLUSION........................................... i9
- 6. REFERENCES********************************************************** 23 APPENDIX A************************************************************** A-1 APPENDIX B************************************************************** B-1 LIST OF FIGURES Figure Title Page 1 Comparison, for Zion, of the airborne aerosol mass suspended in containment with and without containment sprays.............. 6 2 Comparison for Surry of the airborne aerosol mass suspended 1n containment with and without containment sprays.............. 7 3 Removal constant versus normalized, new drop surface area; fresh spray data................................................ 17 LIST OF TABLES Table Title Page 1 Sunmary of CSE First Spray Results ****************************** 10 2 Sunmary of PISCO First Spray kesults **************************** 12 3 Sunmary Results of JAERJ Tests ********************************** 12 A-3
- 1. INTRODUCTION Conmercial pressurized water reactors (PWR) are equipped with Containment Spray Systems (CSS) to limit the peak pressure in containment below the design pressure in the event of a blow down associated with 1 design basis accident.
The CSS is composed of spray pumps, spray rings, nozzles, and necessary pipes and valves. Coolant ts supplied by the Refueling Water Storage Tank. Many CSS are also equipped with a spray additive tank and an associated pump.
These additives are intended to increase the capacity of the coolant to absorb gaseous iodine tn the event of fuel clad failure or core 11elt. Hence, an f111portant secondary function of the CSS ts the attenuation of fission product~
released to the containment. Section 6.2.2 of the Standard Review Plan ($RP) describes the performance objectives of the CSS as a heat removal system.
while SRP 6.5.2 addresses the function of the CSS as a fission product cleanup system.
The performance objective of CSS as a fission product cleanup system, given the source term assumptions given tn Regulatory Guide 1.4, 2 coupled with the containment leakage rate is ti aid tn meeting the design basis accident (DBA) dose guideline of 10CFRlOO.. The basts of source term prescript;on given in Regulatory Guide 1.4 is given in Reference 3. The source term con-sists of lOOi of the noble gases (Xe, Kr), 2si of the iodine and li of other solids. Jodi ne 1s assumed to be primarily gaseous based on the observed release from the Windscale acc;dent. (Regulatory Guide 1.4 further prescribes the following iodine chemical composition: 911 elemental, Si particulate, and 4: organically bound). The Atomic Energy Commission adopted this source term to establish criteria for licensing of plant/site combinations and assessing the potential hazard to the public. The use of this hypothetical source term, which would be associated with substantial core damage to assess the conse-quences of a DBA, which should not result in any substantial core damage since the Engineered Safety Features (ESFs) should terminate the accident prior to core damage, was viewed as intentionally conservattve. Application of this approach led to the conclusion that gaseous iodine dominated the off-site radiation doses. This tn turn led to increased efforts to scrub iodine from the containment atmosp~ere, including the use of a chemical additive to increase the effectiveness of the sprays to absorb and retain gaseous iodine.
On March 28, 1979, Three Mile Island Unit 2 experienced a partially
- ittgated loss of coolant accident. Substantial core damage occurred and significant amounts of radionuclides were released from the fuel.
Environmental monitoring of the accidental releases indicated 0.02 to 0.08 of the noble gas inventory and only 3 x 10.. 1 of the iodine inventory were released to the environment. No IR!tallic radionuclides are known to have been released. The difference in the fractional releases of noble gases and iodine were attributed to the following:
- 1. Noble gases are inert, volatile, and only slightly to fairly soluble tn water.
- 2. The chemically reducing environment in the reactor vessel promotes the stability of cesium iodide which ts nonvolatile (in the containment atmosphere) and water soluble.
- 3. Injection of sodium hydroxide into the CSS would have enhanced the absorption of gaseous iodine if 1t was released during the accident.
- 4. Filters effectively trapped iodine in the 1uxf1 hry fuel" handling building from which environmental releases occurred.
The inference that the majority of iodine released from the TMI*2 reactor vessel was CsI and not molecular iodine focused attention upon the TID-14844~
source term usumptions and the measures taken fn response to these assump-tions, e.g., the design of the engineered safety features.
This report focuses on the technical data base that fs 1v1f l1ble to sup-port the use of chemical 1dditf ves fn the CSS. Computer searches of several literature data bases were al so carried out to identify relevant materials.
These searches are documented f n Appendix A; A-5
- 2. PAST AND CURRENT SOURCE TE~M CHARACTERISTICS
- Regulatory Guide 1.4 prescribes that the following source term be con-sidered, for example, in assessing the D~A doses guidelines values set forth in 10CFRlOO:
- 1. 100~ of the noble gas inventory,
- 2. 251 of the iodine inventory; Composition
- 911 molecular, 51 particu-late, 41 organic.
- 3. Release to the containment assumed to be instantaneous and well mixed in the containment 1tmosphere.
NUREG-2772 addressed the impact of source term assumptions on regulation. Specifically, the impact of the observation that particulate Csl and not gaseous I 2 was the predominant chem1 cal form of iodine rel eased to containment was assessed. Based upon the analyses presented in NUREG-0772, it would appear that except for those accidents in which the fission products are released through water, the amount of iodine calculated to be released would not be substantially reduced by the chemical form (1 2 or CsI).
As a result of the observations of radionuclides released at TMI, sub-stantial research efforts were initiated. An effort sponsored by the NRC has resulted in a set of computer codes, the Source Term Code Package (STCP),
which simulates the progression of severe nuclear reactor accidents and esti-mates the release of materials from the fuel, through the reactor coolant sys-tem and to the environment. The function and status of these codes are described in NUREG-0956: 6 some typical results of the code package and pre-1imi nary observations regarding accident source terms are also presented. In addition, the basis of STCP me;hodology has been reviewed by a study group of the American Physical Society. While this review understandably noted ma!'ly areas of uncertainty and identified several phenomena not fully analyzed, it generally concluded that considerable pro~ress has been made since the publi-cation of the Reactor Safety Study (RSS). (The RSS, published in 1975 1 also predicted substantial releases of gaseous iodine to the containment, and was more or less consistent with the Regulatory Guide 1.4 Source Term Prescript 1on.)
In c0111parison to the Regulatory Guide 1.4 source terms, several substan*
t;a1 d;fferences exist in regard to the character15tics of the fission product (FP) release predicted by current. state-of-the art methods of source term estimation. These characteristics have been described else.,,here, e;g., see References 7 and 8. A brief sunmary of the characteristics is given below.
The total release of FP material to the containment can be divided into two distinct phases: the initial, in-vessel phase where 1R1terial is released from the damaged or 11el ted fuel and the ex-vessel phase where *ateri al *is released from the core/concrete interaction. In the in-vessel phase, the release is dominated by noble gases (*1001 release), cesium (*100%),
iodine (*100\), and tellurium (*30-70'1), which are rather volatile at the temperature excursions predicted duriny core deyradation. Very much smaller 1mounts of the refractory groups, Ba, Sr, Ru, La, and Ce, are predicted to be released in-vessel. As the volatile 1n1terials, with the exception of the
noble gases, migrate away from the core to cooler regions in the reactor cool-ant system (RCS), they are assumed to condense on surfaces or onto aerosols.
8aseg upon observations 111ade at TMl-2, and subsequent thennochemital analy-ses, iodine 1s assumed in the STCP to be present as Csl, and any iodine release from the RCS is 110deled as an emission of CsJ in aerosol form. The behavior of aerosols released from the RCS can be modeled in several alternate fashions which depend upon the particular type of reactor and sequence being considered. For fWR sequences, where the RCS blowdown 1s vented through a suppression pool, aerosol decontamination factors (OF) are calculated by SPARC, a computer code within the STCP. Similarly, for PWR's~equipped with ice condensers, the ICEDF code within STCP is executed to estimate aerosol OF's 1ssoc1ated with this ESF. Ultimately, the behavior of aerosols 1n the containment atmosphere 1s simulated by the code NAUA-4. This code 110de1s several natural processes, e.g. aerosol agglomeration and settling, that can deposit airborne aerosols onto reactor surfaces 1nd, hence, result in a decrease in the airborne activity. The duration of the in-vessel FP release is limited to the period from the start of core degradation to bottom head failure. The onset of core damage is plant and sequence dependent. Typical estimates of the times for the start of core melt ire 25 111nutes 1 Surry AB sequence, and 135 minutes, Surry SzD, 8 from the time of scram. Jn the AB sequence, the CSS is assumed to fail. In the 520 sequence, the CSS is opera-tional, and sprays initiate 20 minutes after scram. It is interesting to note that by the time the release of FP's has begun, the CSS has already entered the recirculation mode. It should be noted that current regulatory guidance effectively requires inrnedhte injection of additives into the sprays, once the CSS is initiated.
The ex-vessel FP releases result from the core/concrete inter1ction where gases generated from the decomposition of concrete sparge through the molten core debris. The large volumes of gases which pass through the melt and the increased surface area 1ssociated with these gases accelerate the vaporization of melt constituents, which subsequently condense into 1erosols after leaving the 111e1t. Another mode of 1erosol generation is also lnOdeled. This fs the formation of mechanical 1erosols which are a result of the gas bubbles break-ing through the upper melt surface. Hence, 111 FP release from the core/con-crete interaction is modeled as being in aerosol form. The ex-vessel release is dominated by the Ba, Sr, Ru, La, and Ce groups. In general, a small frac-tion of the core inventory of iodine is predicted to be retained in the core debris at the time of bottom head failure and is passed to VANESA, a computer code in the STCP which estimates ex-vessel releases, for release during the ex-vessel phase. VANESA assumes this iodine release to be in the form of CsI aerosol. The duration of the ex-vessel release starts shortly after bottom head failure 1nd is typically calculated ten hours beyond initiation, although the majority (-90i) of ex-vessel release generally occurs within three hours of the initiation of the core/concrete'1nteract1on.
Jn sunmary, results of severe accident simulation with the state-of-the-art 111thodology incorporated into the STCP indicate two phases for fission product release. The in-vessel phase 1s associated with core degradation and releases are dominated by noble gases, cesium, iodine, and tellurium. Iodine 1s assumed to be in the chemical form Csl. With the exception of the noble gases, all releases from the RCS are in 1erosol form. The ex-vessel release phase results from the interaction of the molten core and the c.,ncrete A-7
-s-basemat. The ex-vessel rel ease 1s dominated by the Ba, Sr, Ru. La. and Ce groups. Small amounts of volatiles are also released ex-vessel. nota~ly iodine 1n the form of CsJ. All releases are in the form of aerosols. Wnen appropriate intermediate codes, SPARC ANO JCFDF, esti*ate FP aerosol retention in ESFs. The removal of ~irborne aerosols, generat*d either in-vessel or ex-vessel. by natural deposition processes 1s estimatezt ira NAUA.
The STCP, as currently implemented, iloes not edel any gaseous iodine release, nor is there any explicit lllDdeling of iaseous iodine behavior, or the effect of any chemical additive to the css. The physical washout of aerosol by sprays is modeled and results in substantial removal of aerosols from the containment atmosphere. Figure 1 shows a typical comparison of the accumula-ted. masses of material leaked to the environment for a sequence, Zion 520-c, where the CSS OP.trates and another sequence, Zion THLB'*£. where the CSS is 1s1umed to fail. 8 A similar comparison for the Surry plant is shown in figure
- 2. The reduction of Csl aerosol leaked to the env1rorment for these reactor-sequence combinatfDTlS is 1pprDai1111tely SO to 10 , respectively.
Hence. given the current best estimate Df the 'fission product, release charac-teristics and the proceses 'tilat. .c.an Kt im 'the 'FP releases, the physical action of the CSS and natUT~1 ~ol '911D¥11 1""0Ce5Ses which proceed with or without the CSS can substantial*.Y -reduce '11rborna ~oncentration of FP's. Th;s includes iodine as it is modeled to awear 1n aerosol form. Jt 1s noteworthy to mention at this point tflat the tsS cftemi~al &dtfitive(s) effect only gaseous forms of iodine, 12 , HI, and, ~ependin9 upon the specific additive, organ;c iodine, and do not contr;bute ta the p~.1jal removal of aerosols. Aerosol removal by CSS washou.t ts awiela:l a .. ..,.urely 4>hy.si~*1 process. The most co:n-111on A,d~t\ye (namel.r. NaOH) 1 1'eweffT'., *Y l)ley e -secondary but important ro1 e in f4i~~ the sump pH and.1nitiptina 1"Ui4J~sis assi£ted evolution of iodfoe tn gaseous form.
A-8
'b
- est _____ .............. -------- CSI CSOH J[ _._._ **********
_________ .,.. ... CSOH TE: .. ..........
'O
- OTH£RS
, 'b OTHCRS *-*----**-- ,.-
,./ *'
l "o
t ii
!/ ! __.,.,..--,-----*-*-
/
"o
- : ,-----* ~--
C9 en
!/ / !I _J . I
>I a::o c_
'° i
- 1 ,
... I ,.l 0
D
' '1 I' I "o
I' I I
"o- I I
"o-.~u-------------..-~-..d_____________..___,*d
-.0' "o*-1-------------.-------P"----.----.---...-.-...-.....-t Tltl:, nJN 10' TJHE:, HIN lCNU.lllTrD ltAkfl -..1DCS, SzD
- c. ~OJQ.,.m, IUlfD llUtllDCS. nu* - *.
Figure 1 C~artson. for Zton, of the ai rbome aerosol IMSS suspended in containment with and without containment sprays.
.. r*t.. .............. ~~est C5lJH -*-*-
,. ... *-l"l ** TE
- o. TC
- OTHE.RS.. *** :::::::1::1*******
- JTHCRS *****-**-**
,................. /
I "b. I "a. '
I I
. I I
- r-**-*-*--*-*-*--~*#f/11,.._.
I I
"o- :.
I
, ~
../
I I
I
- -------------------.,-~--------
I I
I
~
- 1--------
I I
I a,. e'b i\
tnO cn cn- en a:
¥! I I: I I
~o a!
~-
I b~-.
t-
. I.
\, . I
\*--*-*-*-*-
'o II "o
- 'o I
T~~ i .
to' 10 1
7* lo' tD' Ttt1';, t1tN
- C<<N.UlD l[lrtlt JIUCllllU. S1'J** *c~ *no l [Ar(ll JIUCll DU * , .... , ***
rigure 2 Co"llarison for Surry of the airhorne aerosol mass suspended in containment with and without cont~inm~nt sprays.
- 3. CSS EFFECTIVENESS AS A FP REMOVAL SYSTEM As mentioned heretofore, adoption of the TID-14844 source term assump-tions to assess the consequences of OBA resulted in the conclusion that molec-ular iod;ne d~ninated off-site dose. This, in turn, fostered consideration of the CSS as 1 fission product removal system and resulted in the addition of additives, namely, sodilJTTI hydroxide for 12 and HI removal and sodium thio-sulfate for the further removal of organic iodine, to increase the effective-ness of the spray solution to absorb and retain iodine.
A primary data base upon which 111.1ch of the thinking regarding the effec-tiveness of spray as a fission product removal system was the large scale con-tainment system experiments (CSE). 10 The CSE were carried ou\ tn a one-fifth lfnear scale containment having an internal volume of 751 111
- Experiments A-3, A-4, A-6, A-7, and A-8 were performed to determine the effectiveness of CSS to remove airborne FP. The major variables considered included conta1n-111ent temperature end pressure, spray solution composition, and initial or recirculatory phase. Details of the experiment may be found tn the original references 10 or the concise summary provided by Albert, 11 which is reproduced here as Appendix B. The results ere displayed tn the appendix in Figures 46 through 50 (original text numbering) as plots of airborne iodine to the con-centration versus time. The absorption iodine was generally interpreted as being governed by a f;rst order process:
de dt * *AC, where C is the airborne concentration of iodine, t is the time, and A ts the first order removal constant. The results are tabulated 1n Table 1. Only the first spray per;od results are given since other processes, such as desorption from wall or the effect of inhomogeneous mixing, complicate the interpretation of subsequent spray per;ods.
Several observations can be noted. The DF's, ratio of iodine concentra-tion prior to spray initiation divided by the concentration tnmediately after spray has been stopped, range from 5 to 100. The differences in run A-3 and run A-4, may in part, reflect the change tn spray flow rate. The differences in A-4 and A*6 could result froin either the increased buffertng capacity of the spray solution or the change in initial containment temperature, 25 versus 124°C, respectively. However, a comparison of runs A-5 end A*6 with run A-7,
- here the solution ts unbuffered H3B03, suggest the latter. In addition, it ts interesting to note that runs A-6 and A-7 gave comparable DF's, indicating that tn the initial injection phase buffering, pH control, had little observ-ible effect. A superficial comparison of the results given in Table 1 for runs A-7 and A-8 indicate comparable performance on the basts of DF, while Sha.ting significant difference tn 1 or tt/2* This results from two changes in the expert mental procedure: spray nozzles that delivered a smal 1er 111ean drop she and a $horter duration of 1n1t1al spray operation. Given these changes to the experimental protocol, the results of run A-8 would appear con-sistent with the former runs.
-10..
Table 1 Summary of CSE First Spray Results 1 Run Solution Compos1tionb t t (min) A (min* 1) DFd 112 A3 525 ppm H3B0 3 , pH 9.5 s.s 0.13 5 A4 525 ppm H3eo 1 , pH 9.5 1.4 o.so 100 A6 3000 ppm H3B0 3 , pH 9.5 2.1 0.33 30 A7 3000 ppm H3B03, pH 5 2.2 0.32 30 AS 3000 ppm H3eo 3, pH 9.5 0.64 1.1 30 1
Adapted from Reference 10.
~Fresh room temperature solutic~.
ccorrected for other removal mechanisms, e.g., reaci1on wall; corrections were <lOi; t112
- ln 2/A.
dRatio of airborne iodine concentration direct before and after spray opera.
ti on.
A*l2
Two additional experimental sources of information on the effectiveness of spray additives were 1ocated. These experiments were performed 1n the PISCO 10 faci1ity 12 and by JAERI. 13 The PISCO 10 11e>del containment had an interval volume of gs*m 3
- Twelve experimental runs were carried out. Service water and 11 sodium thiosulfate (Na2S203) solutions were tested. Service water can certainly be considered additive free. The authors of Reference 12 concluded that the re~oval rates were similar for both service water and 11 thiosulfate solution. A summary of these results is given in Table 2.
Nishio et 11. 13 quoted results obtained at the Japan Atomic Energy Research Institute (JAERI). ~ver, no citation was given and the original manuscript could not be 1oc1ted. Based upon the description given by Nishio, the experiments were carried out in a 708 m3 cylindrical vessel. Two experi*
ments, BIS*l and BIS-2, were done under conditions that simulated a BWR LOCA.
One experiment (PIS-9) was performed for conditions which simulated a PWR LOCA. Summary results are given in Table 3. The range of values observed in the JAERI tests is in fair agreement with previous studies. However, direct comparison in all cases is not possible since the spray flow rates are atypi-cal of the regime anticipated in domestic PWRs. However, these results are applicable to BWR contai nmeF1ts .*
A*13
-12*
Table 2 Summary of PJSCO First Spray Results*
Solution Compositionb A (min- 1)
Run t 112 (min) 101 11 Na 2S20 3, 62°c 1.3 0.53 106 lS Na 2S203 , 3o*c 1.0 0.69 107 IS Na 2S20 ! , 1a*c 0.3 2.3 108 lS Na 2S20 3 , 14*c 1.5 0.46 103 service water, e6°c 3.0 0.23 102 service water, eo*c 0.3 2.3 104 service water, 16°c 2.0 0.35 109 service water, 26°c 3.0 0.23 110 service water, 34*c 2.0 0.35 111 service water, 3S°C 3.5 0.20 112 service water, 33°C 1.2 0.58 1
Adapted from Reference 12. DF's not calculated, since spray duration *var;ed widely.
bsolution composition, temperature of spray.
Table 3 Summary Results of JAERJ Tests*
Run Solution Composit;onb t 112 (min) A (min'" 1)
BJs .. 1 pure water, 1o*c 2.9 0.24 815*2 pure water, 1o*c 6.5 0.11 PJs ... g 1.4S H,eo,. pH 9.6, 40°c o.so 1.4
- Adapted from Reference 13. DF not calculated; insufficient data.
Solution composition, temperature of supplied spray.
A-14
- 4. SPRAY MODELS 4.1 Analytical Procedure The original treatment of the removal of iodine from the containment atmosphere is presented in Reference 10. Thf sunmary description given here is based upon the discussion given by Albert. 1 The removal of iodine from the containment atmosphere 1s traditionally eodeled as a first order process:
dC * *AC, (1)
Tt where C is the airborne iodine concentration, t 1s the time, and 1 is the removal rate constant. Integration of Equation (1) yields with c0 is the initial concentration of iodine. The removal coefficient has been defined 10
- 1 ~ as A
- FPt/V, (3) where F is the volumetric flow rate, P is the partition coefficient of iodine between the spray liquid and the gas phase, V is the sprayed containment vol*
ume, and t is the removal efficiency. The removal efficiency has been theo*
retically defined for several mass transport limiting processes. These expressions are for the stagnant drop model (4) for the stagnant film model (S) for the well mixed drop model c
- 1
- exp [(*6 t t )/dP), (6) 9 1 A*15
end for the gas phase controlliny resistance model (7)
In these equations, kg
- gas film mass transfer coefficient, kt
- liquid film mass transfer coefficient (no reaction),
t1 ** drop exposure time, d
- drop diameter, o1
- liquid phase diffusion coeff;cient, t
- 4 0 t /d 2 ,
1 1
- n
- nth root of the equation an cot(an) + Nsh - 1
- 0 h
- drop fall he;ght, and vg
- terminal drop velocity.
In the initial spray phase, when the solution is fresh and therefore con-tains no dissolved iodine, 1t might be expected that the mass transfer to the drop is limited by gas phase transfer. Combining Equations 3 and 7 one obtains A* (8)
Note that 1f 9as-phase transfer 1s the rate limiting step, then the first order removal coeff;cient 1s predicted to be independent of solution composition. Similar first spray removal rate coefficients for various so,u-tions have been observed (see previous section), although the equilibrium par-tition coeff;cient certainly does vary with the composition of the solut;on. 15 Another variable, which will f;nd use in the next section, is the total drop surface area, A, created per unit time per unit sprayed containment volume.
A-16
-1s-(9)
A is simply the total surface area (S1
- nSd) created by drops of mean diameter, d, at the flow rate, F, and divided by V, the sprayed containment volume. Substituting Equation 9 into 8 yields:
>.
- kg h A/v g * * (10) 4.2 Reevaluation of Existing Data Jn an earlier section, available data on the iodine removal effectiveness was reviewed. Care was taken to select and display only data for fresh spray solutions. In both the CSE and PISCO experiments, multiple spray periods, including recirculation, were investigated. These latter results are expected to reflect the effects of increased iodine concentration in the drop, as well as the heretofore mentioned complications of wall desorption and inhomogeneous mixing. However, when the spray solution was fresh, all solutions appeared to effectively reduce the airborne iodine concentration, regardless of the pres-ence or absence of an active spray additive. To demonstrate this contention, and account for the major variables causing variation, selected first spray data from the CSE, PISCO, and JAERI tests are displayed in Figure 3. In this figure, the first order removal rate coefficient 1s plotted against the nor-ult zed total new drop surface, A. Also displayed are the range of fl ow regimes of several typical 16 PWR's. These were estimated from information obtained from the Surry FSAR and information provided in References 17 and
- 18. The upper limit of a range represents both spray header systems are oper-ating, while the lower limit represents operation of only one of the two redundant spray systems. Approximately 3/4 of the PISCO data and both JAERI BWR test data are not plotted, as they would lie well beyond the anticipated range of A for domestic PWRs.
Although the plotted data exhibit some scatter, a generally good correla-tion is found. Hence, when spray solution is fresh, the re1noval of iodine from the containment atmosphere is dominated by gas phase mass transport and is effectively independent of the equilibrium iodine partition coefficient of the solution, and primarily controlled by the amount of available surface to which iodine may be transported. At a first level of approximation, the good correlation of >. and A observed indicates that the combination of terms not explicitly examined in Equation 10 1s effectively constant or slowly varying over the range of experimental conditions investigated. It should also be noted that 3/4 of the PISCO and JAERI data not plotted. that data which was taken in a fl ow regime atypical of domestic PWR 's, does not correlate well with A al one. There are several potential reasons why this occurs, however, these have not been examined since it is felt that the experimental conditions are basically atypical of those conditions associated with domestic conmercial PWR's.
A-17
- 16-Va1ues of kg, computed from Equation (10) and for the experimental data disp1ayed in Figure 3, are given 1n Table 4. For comparison, estimates of kg and 1, based upon a well known correlation for heat transfer to a si ng1 e dtop, are also given. ~$um1ng the minimum observed experimental k , 3
- /min, maximum fall height ll, 18 and that both spray headers are opera~fog, first order removal coefficients of 0.8 m;n-1 and 2.0 min*l are estimated froni Equation (10) for the Surry and Zion plants, respectively. If it is 1ssumed that structures, e.g., the reactor pressure vessel and steam gener-ators reduce the effective drop fall height by 50% to 601 of the maximu", then estimated 1 's of 0.4 mi n*l and 1.3 mi n*l are obtained for these phnts, respect 1vely.
A-18
100,.....-----------------------------------------------.
50 *
~
..c l 5
I Surry 11 Zion I e CSE
- PISCO
- JAERJ l
0 l 2 3 4 5 6 A (m--minr 1 Figure 3 Removal constant versus normalized, new drop surface area; fresh spray data.
S. DISCUSSION AND CONCLUSION In previous sections of this report, the current best estimate descrip-tion of source tenn characteristics was sunrnarized and available data on the effectiveness of spray as 1 fission product removal system were presented, reviewed, and reevaluated. CSS can effectively reduce airborne concentrations of aerosols which current methods of source term estimation predict to over-whelmingly dominate fission product releases. The inert noble gas releases are unaffected by sprays. Aerosol removal by the sprays is a physical process and this process 1s not altered or aided by the presence of chem;cal additives. Other natural processes, modeled by NAUA-4, can also reduce air-borne aerosol concentration. The relative benefit accrued is closely related with the time available prior to containment f1ilure for these natural proc-esses to fct. Date presented in Chapter 2 indicated that the combined effect of these 1rocesses can reduce airborne CsI 1erosol concentrations by a factor 50 to 10 , depending upon the specific reactor and accident sequence being exam;ned.
Current NRC sponsored anaiyticai modeling of severe accidents, i.e., the STCP, does not predict the emission of gaseous iodine in the anticipated acci-dent environment associated with conrnercial LWR's. This is based upon an examination of the TMI-2 accident, and subsequent thermochemical analyses.
The actual chemical form of iodine is still subject to a degree on uncertainty. Evolving experimental and analytical evidence 7
- 19
- 2 ~ indi-cates boron may be chemically associated with Cs; hence, boron may be in com-petition with iodine and potenti1lly liberate iodine in another form, possibly gaseous. Preliminary experimental results 21
- 22 also suggest react;on of CsOH and Csl with the stainless steel surfaces of the RCS, w1th the reported em;ssion of gaseous iodine in some cases. The reproducibility of these exper-iments are currently being investigated. In addition, some experimental evi-dence23 has been obtained that indicates the conversion of Csl aerosol to gaseous iodine during hydrogen burns. These observations certa;nly reinforce the diversity of material interactions and phenomena that can occur and g;ve rise to* uncertainty. Additional research 1s in progress and 1s required to resolve this uncertainty. However, regardless of the extent of gaseous iodine conversion, the washout by an operational CSS would occur and the data pre-sented in Chapters 3 and 4 clearly suggest that it is effective. Moreover, the fresh spray d1ta suggest efficient iodine removal regardless of the pres-ence of additives during the initial injection phase. This is not to sa1 that NaOH 1s not ultimately required to increase the absorption capacity of the spray solution and mitigate iodine reevolution fron the reactor sump. On the contrary, sufficient evidence exists to warrant pH control in the long term.
The regulatory option to be reconsidered 1s whether or not the presence of NaOH is required during the initial injection phase of CSS operation. Two items affect this decision. One, is the effectiveness of fresh spray solution, end the other is that sfnce the CSS is activated on high containment pressure, it is quite possfble that the CSS will have switched from the injec-tion to the recirculation phase prior to the release of any fission product activity.
A potential 1ltern1tive is to add pH control directly to the reactor sump rather than in the initial injection supply of the css. Additionally, it would be attractive to initiate pH control on some feedback directly related to the release of activity rather than on high containment pressure. This
-20 ..
would have the obvious advantage of not 1ntroduting the additive(s) until required. A setondary benefit should the reattor intident be terminated with-out the release of FP attivity, would be 1 simplified cleanup retovery.
A-21
Table 4 Comparison of Experimental and Estimated Mass Transfer Constanu m
Run A(m-min)-1 te(m;n) >.(min* 1) 1 est kg mm kg est A3 0.40 1 o.osob 0.13 0.14c 6.5d 6.7e A4 1.55 o.oso 0.50 0.52 6.5 6.7 A6 1.55 0.050 0.33 0.65 4.3 8.4 A7 1.55 0.050 0.32 0.65 4.2 8.4 AS 2.45 0.083 1.1 1.8 5.4 8.9 PlS-9 5.36 0.092 1.4 3.7 2.9 7.4 109 0.74 0.029 0.23 0.15 11 6.7 110 1.35 0.020 0.35 0.18 13 6.9 111 1.19 0.020 0.20 0.16 8.6 6.9 112 0.79 0.029 0.58 0.16 25 6.9
(*)calculated from reported data for F, V and d.
(b)te
- h/v 9
- Velocity calculated from Vg * (4(Pa
- Pg) g d/3 Pg t)l/2, where t
- 18.5/Re 0
- 6 , and Pu is the drop density, Pg is the gas density, g is the gravitation constant, t 1s the drag coefficient and Re is the Reynolds numberi Handbook of Multiphase Systems, G. Hetsroni, Hemisphere Pub. Corp **
New York, NY, 1982.
(c)Aest
- A te kg est*
(d)t 9 back calculated from experimental data.
(e)kg est estimated from the correlation of Ranz and Marshall, Chem.
Eng. Prog., 48, 1952.
kg
- D/d (2+0.6(pdvg/~)l/2(~/pO)l/3)~
where O fs the diffusttfvtty of 12. All gas phase variables are for air.
A-22
- 6. REFERENCES
- 1. u. s. NRC, *standard Review Plan for the Review of Safety Analysis Reports for Nuclear Power Plants: NUREG-0800, July 1981.
- 2. U. s. NRC Regulatory Guide 1.4, *Assumptions Used for Evaluating the Po-tent1a1 Radiological Consequences of a Loss of Coolant Accident for Pres-surized Water Reactors,* Rev. 2, June 1984.
- 3. U. S. NRC, 10CFRlOO, *Reattor Site Criteria,**27FR3S09, April 12, 1962.
- 4. J. J. DiNinno et al., *calculation of Distance Factors for Power and T~st Reactor Sites,* u. s. Atomic Energy Co11111ission, TID-14844, March 1962.
- 5. u. s. NRC,
- Technical lases for Estimating Fission Product Behavior Dur-ing LWR Accidents,* NUREG-0772, June 1981.
- 6. M. Silberberg et al ** "Reassessment of Technical Bases for Estimating Source Terms," NUREG-0956, 1986.
- 7. R. Wilson et al., "Report to the American Physical Society of the Study Group of Radionuclide Rel ease from Severe Accidents at Nuclear Power Plants," Rev. Mod. Phys., 57(3), Part 11, July 1985.
- 8. J. A. Gieseke et al., "Radionuclide Release Under Specific LWR Accident Conditions," BMI-2104, July 1983.
- 9. H.P. Nourbakhsh et al., "Effectiveness of BWR Pressure Suppression Pools 1n Retaining Fission Products,* Brookhaven National Laboratory, Technical Report A-3788, July 1, 1986.
- 10. R. K. Hilliard et al., *Removal of Iodine and Particles from Containment Atmosphere by Sprays: Containment Systems Experiment Interim Report,"
BNWL-1244, February 1970.
- 11. M. F. Albert, "The Absorption of Gaseous Iodine by Water Droplets,"
NUREG/CR-4081, July 19!5.
- 12. S. Barsal i et al., *Removal of Iodine by Sprays in the PISCO 10 Model Containment Vessel," Nuc. Tech., 23, August 1974.
- 13. 6. Nishio et al., *containment Spray Model for Predicting Radioiodine Removal in Light Water Reactors,* Nuc. Tech., 54, July 1Y81.
- 14. A. K. Postma et al., "Technological Bases for Models of Spray washout of Airborne Contaminants in Containment Vessels: NUREG/CR-0009, October 1978.
- 15. Reference 14 and citation' contained therein.
- 16. VEPCO, Surry Power Station Units 1 and 2, Final Safety Analysis Report, Virginia Electric and Power Company.
A... 23
.24..
- 17. *westinghouse Nuclear Training .... four Loop Plant Information Book, 11 West ..
1nghouse Electric Corp., P;ttsburgh, PA, 1978.
- 18. U.S. NRC, *Preliminary Assessment of Core Melt Accidents et the Zion and Indian Point Nuclear Power Plants and Strategies for Mitigating Their Ef ..
fects,* NUREG-0850, Nov. 1981.
- 19. D. Powers, *High Temperature Fission Product Chemistry,* Severe Fuel Dam-age end Source Tenn Research--Program Review Meeting, Oak Ridge, TN, April 1986.
- 20. A. T. O. Butland et al.,
- The Effect of Variations 1n Chemical Species end Associated Properties on Primary System Retention in PWR Severe Acci-dent,* to be published.
- 21. R. M. Elrick et al., *Reaction Between Some Cesium.. Jodine Compounds and the Reactor Materials 304 Stainless Steel, lnconel 600 and Silver,"
NUREG/CR-3197 (3 volumes), June 1984.
- 22. o. Powers and R. Elrick, *sNL Radiation Effects Experiments,* Severe Fuel Damage and Source Term Research--Program Review Meeting, Oak Ridge, TN, Apri 1 1986.
- 23. L. S. Nelson et al., "The Behavior of Reactor Core-Simulant Aerosols During Hydrogen/Air Combustion," Thirteenth Water Reactor Safety Research Information Meeting, Gaithersburg, HD, Oct. 1985.
A.. 24
A*l APPENDIX A Computerized searches of the literature were performed to obtain perti-nent citations on the subject containment spray and spray additives for PWRs.
The following databases were queried:
EDB (Energy Data Base, DOE/RECON, 1976-present)
The search resulted in 171 citations; of which 18 were selected and ac-quired in support of the research project: and NSA (Nuclear Science Abstracts, DOE/RECON, 1967-1976)
The search resulted in 8 citations.
The search strategy was:
PWR and
[containment spray systems or (containment system) and (atmospheres) or
((sprays or droplets or particles or iodine removal or particulates or additive(s)) and (containment systeMs))].
A-25
APPENDIX B The following chapter was reproduced from NUREG/CR-4081. It provides a brief deser1pt1on of the CSE and a concise sunnary of the CSE results and their precision.
A-26
BS
- 7. EXPERI~ESTAL DATA The experimental data used to compare with the results of the spray
- odel are froa the Containment System Experiments (CSE).3 Experimental runs A-3, A-4, A-6, A-7, and A-8 of this series are large scale spray ay1te111 tests to detenaine the effecti\*eness of a spray aystem for
- removing airborne fi11ion products. The results of these tests are reported in terms of the 1as phase elemental iodine concentration versus time and also in terms of the liquid phase elemental iodine concentra-tion wraus time. The parameters for the apray experiments are the apray flux, the *rop size, the 1as phase temperature, pressure, and humidity, and the liquid spray composition. The physical dimensions of the CSE vessel are listed in Table 3 and are shown in Figure 44. Since Table 3. Physical conditions common to all spray experiments (Hillard3)
Volume above deck including drywell 21,005 f t3 595 mw1 Surf ace area above deck including 6,140 ft2 569 m2 drywell Surf ace area/volume o. 293 ft-1 0.958 m-1 Cross section area, main vessel 490 ft2 45.S 1112 Volume, middle room 2,089 ft3 59 m3 Surf ace area, middle room 1,363 ft2 127 1112 Volume, lower room 3,384 ft 3 96 a3 Surf ace area, lower room 2,057 ft2 191 112 Total volUD1e of all rooms 26,477 ft3 751 m3 Total surf ace area, all rooms 9,560 ft2 888 .2 Drop fall height to deck 33.8 ft 10.3 fD Drop fall height to drywell bottom 50.5 ft 1s.4 11 Surf ace coating All interior surfaces coated with phenolic paint.a Thermal insulation All exterior surfaces covered with 1-in.
fiberglass *ln1ulation.b 4 Two coats Phenoline 302 over one coat Phenoline 300 pricer. The Carboline Co., St. Louis, :Ussouri.
bk* 0.027 Btu/(hr) (ft 2 ) c*r/ft) at 2oo*r, Type PF-615, Owens-Corning Fiberglass Corp.
A-27
86 flUN RUN Al A*.1.1.1
.. ~~S*lh JILANVIEWOF NOZZLE ARRANCEMENT
,,AIN CONTAINMENT ORYWELL VESSEL LID
...._DROP COLLECTOR C*I SOLUTION
- STORAGE TANK RECIRCULATION l PUMP Fig. 44. Schematic diagram of containment arrange~ent used in CSE spray tests (Hillard3).
these tests were aade in realistic and not idealized equipment and con-ditions, the liquid and gas flow patterns are complex and not well characterized. The results from the new spray model will be cot:1pared vith these results, but no better than approximate agreement can be ex-pected. This data, however, can still provide a *ans for useful and meaningful evaluation of the spray model.
The CSE vessei is a large acale vessel bee Table 3 and Figure 44). The overall dimensions of the vessel are 20.34 meters high and a diameter of 7.62 *ten. The *e11el has a drop fall height of 15.4 aeters. The overall volume of the vessel is 751 cubic meters.
The tests varied the temperature, pressure, pH of the drop, spray nozzle configuration and drop size. The conditions for run A-3 are a temperature of 298K, 1 atmosphere of pressure, pH of 9.5 and a drop dia-.
aeur of 1210 aicrons. For all of the tests, the spray aolution tem-*
perature vas ai 25*c, and the solutions were all buffered. For run A-4, the conditions were the same as for A-3 except for a hi1:her tpray flow rate and a different apray nozzle configuration. Run A-6 increased the temperature of the cas to 397K and the pressure to l atmospheres. Run A-7 changed the pH to 5, lowered the tet:1perature to 39'K and raised the A-28
87 pressure to 3.4 at::iospheres. Run A-8 changed the drop diameter to 770 microns. See Figure 44 for apray nozzle arrangements, Table 4 for spray nozzles used, Table 5 for the atmospheric conditions, Table 6 for the spray flow rates and solutions used in the tests and Table 7 for the timing of the spray periods.
The exp~rimental procedure for the molecular iodine spray absorp-tion tests involved first heating the containment vessel vi th a team until the specified temperature was reached. A flask containing.
aolecular iodine traced *with 1 curie of iodine-131 was heated electri-
- cally. Air was passed over the flask to release molecular iodine.
Samples were taken prior to turning on the sprays to determine how aolecular iodine beh~ves without sprays. After the first spray period Table 4. Nozzles used in CSE spray experiments (Hillard3)
Runs A3, 4 1 6, 7 Nozzle type: Spraying Systems Co. 3/4 - 7C3 Nozzle characteristics: Tog type, full cone A4, 6, 7 Ni.miber 3 12
, ., '°)
':'I"-~
Layout Spacing Triangular 10 ft 5 in
- apart Square gride 6 ft apart W--
.\, . ~ .. .$1 ~
\~~-
Pressure Rated flow MMD 40 psid 4 gpm 1210 ...
40 psid 4 gpm 1210 ...
0 g 1.s 1. 5 Run AB Nozzle type: Spray Systems Co. 3/8 A 20 Nozzle characteristics: Fine atomization, hollow cone Number used 12 Layout Square grid Spacing
~
6 ft apart Pressure 40 paid
~. Rated flow 4 IPID MMD 770 ...
a, 1.s A-29
88 table 5. Ato1115pheric conditions in CSE spray experi*ents (H1llard3) llun lun lun ltun ltun Al A4 A6 A7 A8 Containment *easel No Ho Yes Yes Yes
. insulated forced air circula- Yes Yea No Ho No t1on 4 Start of lit apray Vapgr temperature, 77 77 255 248.7 250 Pressure, plia 14.6 14.6 44.2 50.0 50. 7 Relative humidity, 70 88 100 100 100 End of. lat. *pray Vapor temperature, 77 77 229 234.5 243
- FD Pressure, psia 14.6 14.6 38.6 44.4 48.2 Start of 2nd spray Vapgr tecperature, 77 77 237 240 243 Pressure, psia 14.6 14.6 40.8 46.0 243 Ind of 2nd spray Vapor temperature, 77 77 202 203 188
- FD .
Pressure, psia 14.6 14.6 29.5 36 34.1 Start of 3rd spray Vapgr temperature, 77 77 233 230 218 Pressure, psia 14.6 14.6 40. 7 41.8 32.2 Start of 4th spray Vapgr temperature, c c c 232 247 Pressure, p1ia c c c 42.4 52.4 End of 4th spray Vapgr temperature, c a a 192 175 Pressure, psia a c 0 32.7 32.4 4
Fan without duct located in bottom of drywell.
2400 f t 3/min d11char1**
bAverage of S thermocouples located at various eleva-tions and radU.
0 Ho fourth spray.
A-30
89 Table 6. Spray flow rates and aolutions used in CSE experi111enu (Hillard3)
Jtun Jtun llun l.un llun Al A4 A6 A7 A8 1st apray Total flow rate. gpm 12.8 49 49 49 50 Volume sprayed, 1al 128 490 490 490 150 Spraying pressure, 40 40 40 40 40 plid Solution a G b c b 2nd spray Total flow rate, apm 12.s 49 50 48.5 50 Volume sprayed, gal 385 1480 1500 1455 1850 Spraying pressure, 40 40 40 40 40 psid Solution a a b c b 3rd spray Total flow rate, gpm 12. 5 42 16 45. 5 47 Volume sprayed, gal 735 1890 960 2730 2820 Spraying pressure, 40 29 4 36.5 36.5 Solution 4th spray d e
- e e Total flow rate, gpm g g g 48.6 50.4 Volume sprayed, gal fl g fl 2428 2520 Spraying pressure, g g g 40 40 psid Solution g g g f f afresh, room temperature. 525 ppm boron as H3B03 in demineralized water. NaOH added to pH of 9.5.
bFresh. room temperature. 3000 ppm boron as HJBOJ in demineralized water. MaOH added to pH of 9.5.
0 rresh, room temperature. 3000 ppm boron as H3B03 in de::Uneralized water. No NaOH added. pH s.
dFresh, room temperature dellineralized vater.
- solution in main vessel SUlllP recirculated. No heat ex-changer used.
!Fresh, room temperature. i Vt% Na2S203. 3000 ppn boron as H3B03 in demineralized vater. MaOK added to pH 9.4.
fNo fourth spray.
A-31
90 table 7. Timing of spray periods (Hillard 3 )
Time after at:a~ of iodine rel...u, mn lun ltun tun ltun ltun A3 A4 A6 A7 A8 First spray Start 40 40.5 30 30 30 Stop 50 so.5 40 40 33 J>uration 10 10 10 10 3 Second *pray Start 140 140 80 80 80 Stop 170 170 110 110 117 Duration 30 30 30 30 37 1bird spray Start 1473 1205 15'6S 1323 200 Stop 1.533 1250 1525 1383 260 Duration 60 45 60 60 60 Fourth spray Start a a a 1443 1350 Stop a a a 1493 1400 Duratior. a 50 .50
" ?Z aNo fourth spray.
of each run was started, many samples were taken from the gas phase, from the liquid in the sump, from the wall trough and from the spray drop collectors. When the first spray period was ended, more sa::ples were taken to determine how *olecular iodine acts. A second, third, and sometimes a fourth spray period were used with aany samples taken from the gas and liquid phases. The gas phase concentrations were determined by Maypack samplers (see Figure 45), and the liquid phase concentrations were determined by *asuring the amount of lo.dine-131 tracer present.
For aore 1nf ormation see. Reference J.
Results of these experimental tests *Te ahl:un 1n Figures 46 through.
55 and Table a. Table 8 shows the aater~al bat.nee of iodine for all ol the experimental runs. lt should be 9Dted in this table thO'lt bett.:een 25 .65% and 57 .58% of the iodine deli"'l"ed to t:he containment vessel is unaccounted for and is assisned to be d~osited on surfaces. Figures 46 through 50 show the concentration of eletne1i~al Iodine in the gas phase as a function ~f time. The data is repoTted in terll\5 of the half life A-32
T.1h It- 11. Ind I n1* ..1h*t I .11 h;1 t.1nr*!I
( 11111.tr*I')
Run A1 Run ,.,. Run A7
""" "" za A~ R11n I.out Inn r.r .... za r.rAWl!I t" C:r Mll!I z" er .... %" CraM Al"rn!IHI (',.nt'r.1tt Inn
~t*rtt.._ 11aterl*I 1n1.on 1nn.nn lnl.50 1110. no 1111.mt 1no.nn 1n1.on um.no 1n1.nn ,,,,,_""
r... n*r*tlon app11r*ta1 2.si J.54 1.11 1.11 n .14 n.u 1.n1t 1.ns 1.0 1.u Inject Ion I lne n.n :n.o* 29.12 2111.99 1.49 1.u 2.n5 2.n1 1.u 1.ttn n.12 n.11 lnJHtlon tine
!!""'PIH n.16 o.1r. n.1.5 n.15 1.n1n 1.n2 ft.12
.\t'cnunte* for 25.n 25.nn ]0.5' '.ln.14 2.0 1.0 1.*n 5.14
- 2. "'"
0.1 l*ere* to cnn-talr111tnt (by dll-ferenc*)
75.15 75.nn 70.91 fl.CJ.A6
'"*" 97.)'9 97.51 I
~
r.ra* Cr11M t,,
<An l .it I """° er-11 t z" er ...
,,,. er .... ~~ -'° O.lhen* to con- 75.75 1no.oo 70.91 ton.on
'"* 14 1nn.oo '1.51 ion.on tS.61 1no.oo Ulftllent
'1.n In tt11u1' poolse (Prior to *l!!con-t11*ln*tlon) 0.12 59.81 ]7.67 51.11 H.97 54.H ]9.Z8 40.26 S4111pln 0.48 o.u 8.87 12. 51 o.5'16 o.57 o.5' 0.60 O.H o.n Pur"e to tuck 1).)2 n.69 n.11 1.n1 n.OA6 n.nt 0.16 n.11 0.11 0.11 fll*c nnt
- I 1u1t IM 5.IJfJ 1. 711 o;. '" 1.10 n.1112n o.u 1.16 1. ]9 1.u '* o;o ACC'OUnted for ';2.22 611.91 U.7) 74.)5 5\.41 56.17 u.1* o.u SS.57 511.u
'"' 1t1rhce1 <*1 21.52 11.n6 Ill. IA 2';.ft5 42.91 4).6] 56.111 51.n
'"*"' 11.H ii If I Hl!!nce) dPercent or *t*rtlnr, .....
~Percent of dell*erl!!d *****
e Inc I oile1 11 pr *J 10 I utl nn 11nd *t *
- cnnden**t *
- 92 1 INLET 2 TEFLON BALL *
'1 SIX SILVEFI PLATED SCREENS lllLVEA MEMBRANE FILTER 11 SCREEN 12 SPFllNG 3 NOSE CONE I CHARCOAL LOADED FIL TEFI 13 IODY
- TEFTON GASKET
' STOP FllNG 1* END CAP
' TWO PARTICLE FIL TEFIS 10 CHARCOAL BED 11 OUTLET Fig. 45. Maypack Sampler.
of iodine, defined as t1/2 * - Ln (1/2)/1 , (65)
- 0.693/A , (66) where A ls the removal rate constant.
The reason the data are in this form ls becau!le the old 1pray aodels (Equations I through 7 of Olaptn I, Section 2) are in terms of the removal rate constant. Figures 51 to 55 1how1 the concentration of iodine in the liquid versus time. As can be 1een in these figures, there is a delay in the response of the increase in the concentration of iodine in the liquid phase.
In these tests there are aany processes for the removal of aolecu-lar 1od1ne*fr01D the gas phase. In these large acale realistic tests, there are painted aurfaces, fton-painted surfaces, insulation, 1pray1, wet valls, and dry vall1. All of these features can contribute to iodine aorption, and heat transfer can also have an effect on the re-aoval rate of aolecular iodine fr01D the 1as phase. 'Therefore, one can only hope to develop an approximate model vhich accounts for the major phenomena involved and considers only the removal by the sprays. If one looks at the drop data, these date are .. difficult to interpret, not only because of 1ar.1pllng inadequacies but because the relative fractions of the various iodine forms and particle siz'!s were changins rapidly with time.*3 A-34
llUNl 11.\111 ,,. **** : ' I II
- to'* 111'*
l 1 Cllt[Nf Al fOUINE
- RllN AJ OllOf' SI II 1110pMUO
", 6!;-t fUMCNTAl ~OIHt - ~UH A4
TtUPl RA TUR!:
118 *IMlt - I
~
OllOl'Stlt; FLOWAAT!:
1710Jf MfAU "49 .......
PAlSSllRC
- 14 6 tt1i1 tEMf'CAATUflE: 1r'F SPRAY AODITIVf: 51S 1ttom ftORON ly, ~ 1.4 min PRCSSIJAf: , .. 6 tllia SPRAY AOOtnV!: 52Sl'I"" BORON 10" . N,.011, 11H ' 9 S N..Olt, 11H 9 5
.-.. ..,- to' E
E z-z
"' ~
0 /FIRST Q lctl I-a:
~
c a:
SJ'AAY
~
z z w
u z
THIRD SPRAY 0 0u l\'VATERI u
w
"':I:
f{ 107 1100 "'"':I:
f{ '°'w
- )>> CL CL f{
w U1
"'u f{ u THIRD PRAY
..!' J" IAfCIRCULATIONI 101
/FIRST SPRAY 107 SECOND SPRAY 1(ft..._...__,_,.__...___..__..A..~~..__...-4+-~"-~_._.._ ........~ to*
0 100 1!>0 100 1400 1500 1600 0 so 100 150 200 150 1105 1150 TIM! f1nonl TIME I""" I Fig. 46. Concentration of ele*ental Iodine Ftg. 47. Concentration of ele11ental Iodine In the 111.1 In rOOt11, run A-3 ( 1111 lard l). tn t1te m<Jfn r00111, run A-4 (llfllardl).
ORNl-DMi ..
- ftJI lfD 1o!t 1~,
- 17min UEMCNTAl IODINf - RUN A1 OROP SIZE: 1710 Jf MMD tUM[NTAl tOOINE - AUN A& FLOW RATE: "*""
On OP Silt_ 111011 MMD tt MPE AA TURE: 7SO"F HOW RATE: -49*nom PRE S$URE : 50 111i1 TU.IP[ nA TUAE: 1so; °F SPAA Y AODITfVE: 3000 tlflM
..E 1o' PRFSSllRE -44 '"
SPRAY AOOIT IVE: JOOO ,,,_ BORON N.tOH. t*ll 9 S BORON AS M31MJ,.pH S
- ? ?.
0 Q
...-c i=
...a:.z ...a:"'z
>I. .., 10> ~ 10-1 w
°'
~
0 z
0 u
TMIRO u
... SPRAY w
w -'/ -:' ._,:;. *_ IAECIAC I
"'J: / '"*
- 540 min f(
...w J:
lJ THIRD SPRAY f(
0 i-- J"101 u IRE ClflCIJlAT IONI ......_
SECOflO sPAAY
/
10' 0 ISO 100 IS~ 1600 16~ 0 100 1300 17.JO 1-400 1450 ISOO TIME lnnnl TIME lniinl Ffg. 48. Concentration of eteMentat todlne Fig. 49. Concentratton of eleMental Iodine In the 11t.1ln ro0111, rnn A-6 (lllllardl). In the main roo11 1 run A-7 (Utl lard l).
IMtNl IJWG IM 11'..1' 1 HI IP'..---...-------------------------~---------- ~--~
\ .... ,,_
I Cl UtUH Al IOOltJE . nuu A8 onor size.
flU'.V RATE:
l l f.IP[ RA TURt :
P!l l SSIJll C no,, uuo SO**n 7!JO'*F 50 '"'*
oi*Nt llftli,.. 1\11 f IO~r--T..-------._._.._._._._.___,.,_....
ru srnAY AODITIVl ron ht.'""*
AfllJ ]111 srRAYS 3000 ....... RORON AS II JnoJ* .... !J s ill1h SPRAY HAS 1 wl~*
E N.:i 1S 10 1 IN ADIJIT1m1
~
q z
...c -~. 10-1
-b---<t e
...z It °'
"t RUNA3 z HJOINt CONCENTRATION
... 10' 0
- MAIN VESSEL sur.fP I
w3 uz t::
c 107 A bRYWEll SU~ I 10.000
.,. ...z Ir CESIUM CONCENTRATI
-.. \0 c
.,.c IL
"'uz 0
C> MAIN VESSEL SUMP ORY\"IEll SUMP IL
~
\II u
u 101 1.000 .,.::>_,
u j' 10~
FIRST SECOND SPRAY THIRD -I!
."t
.-:., ... ___ THIRD SPRAY I, t
SPHAY tnO w SPRAY /FIRST 100 ~
/. IRCCIRC t, _,
THtO-
- SPRAY UOUID VOLUMEJ 0
SULPHATE --UAIN YESSE >
I 11H9S 101 ...__._____.____.___....___..._..______._____a.;Po\--'-----"----'
- * - DRYWCLL f0* 1 JO 0 IO I 17 I !JO 180 Ol(JO 120 ~o JJ 0 60
'"°t-*I TIME 103 TIMC '"'"
Fig. SO. Concentration of ete11ental todtnc Fir,. SI. Liquid votu** and concentration In the *ain *roo11, run A-8 ( 1111 tardl). Jn vessel su*ps, run A-1 (lllttnrdl).
ORNl OWGll l\JOt:fD AUNA& CY OW 10** IODINE 0 e
ccs*~* a
- VOLUME - - -
THIRD SPllAY ~
10" .~
-~ 0 Q'
6 10.000 ..J
!!- 10' z
- a z z
- a-I 0
- ...Q c(
w ~
CX> ...z a: ...za: 104'
"" 1.000 ~ "'uz ----i
t Y.
'°' :;)
0u u - UI
..J
""~
0 u
ci .,. --------'--t ~--
w MAIN z / .I 3
101 RUNA4 IODINE VESSEL DRYWEll
- TOO "':J
~
..J I
I I
10' 0
0 4 :J CESIUM 0 I
LFIMT VOLUME
--- 0 0
..J SPRAY to" 1;ins 17!.0 0 100 1~0 100 1565 1G75 0 t.O no 180 7110 TIME Im"
TIME fnn11I Fig. Sl. Lfqutd wlu11e* and concentrat ton Fig. s2. Liquid volumes and concentrat ton ln vl.'sse l snmrs, run A-6 (II t t 1ard l) **
In vessc I . sumps, run A-4 (Ill llardl)
UHNl
- IMlo IM V..JI I fO U
lfllJINC
~lllM
.f.~
n II
~RVWHl e
VOlllME - * - *-***
RUNA8 tOOfNI:: 0 e llr' CESIUM D
- VOLUME - - -
NOTE : 6000 liftofl FROM CV SUMP WAS DISCARDED z
- 1o>
0 j::; g of 6
w
....a: ~
z 10J l! z
- J90 I u 101 I tot 0
j::;
~!?
,r - .
of i *-' .., I /';_: ..
. ____ *' fz 10' - to' 3 r .....-
~
1 w
u
---"'1* -. z 0 I /
u**' [ J--*vi>+*** ~
u u
I, , ., ,,,, ., . .I ,. ,*
I THIRD , ,,,. /
-. f---.. I
1 I I 10 SPRAY I
IRECIRC I *' I ~-
I ' I 'II I
... J....__.__._._...._...._.__~~ .. '.....
I t
I 0 too 1]00 IJ!JO 1450 0 50 100 150 700 750 1350 1400 TIME fmml TIME fminf Fig. 54. Liquid YOlu11e1 and concentration Fig. 55. Liquid YOlU11e1 and concentration In *es set sm*e** run A-7 (1111 iard l). In vessel su*p*, run A-8 (Hlllardl).
98 The data for the gas phase are the result of the combined effects of all of the processes for the removal of molecular iodine from the gas phase. lut if .all 9f the processes except for the sprays exerted only a amall overall effect in the removal of molecular iodine from the gas phase, then these data would be acceptable from the standpoint of use-fulness in determining the efficiency of the spray model. The data for liquid in tl:ae *umps 1hould eliminate some of the sources of error be-cause these data shows how much molecular iodine is transferred to the Uquid. Nevertheless, any iodine that is on the surf aces and is not chemically .held to the walls could be washed off in the 1umps. Since the 1prays were not *tarted at the instant the molecular iodine wa:;
r~leased, a significant amount of molecular iodine releasetS into the containment may have deposited on t~e 1urfaces, and subsequently* been washed off into the sumps or aight have been transferred back into the gas phase later wh~n the partial pressure of molecular iodine in the gas phase was *=aller than the partial pressure of molecular iodine on the surfaces. The latter effect could result in an underestimate of the removal rate of molecular iodine.
To remove some of the possible sources of error, the comparison of experimental results to the results of the spray model will be limited to the area of the drywell. The drywell had a cross sectional area of e.s square meters (which 1s a diameter of 3.35 meters), a drop fall height of 15.4 meters, and a volume of 135.52 cubic meters. For example, in ruri A-3, at the start of the first stray period, the initial 1as phase concentration was approximately 5
- 10 microsrams/cubic,meter (1.97
- 10-7 moles/liter) and the final concentration was approximately 1.25
- 10~ microsrams/cubic meter (4.92
- 10-8 moles/liter). 1lle amount of iodine removed from the gas phase during the first spray was 5.082 grams. Also, at the start of the first spray, the concentration of the liquid in the drywell sump was approximately 8
- 102 micrograms/liter and the initial volume was approximately 150 liters. At the end of the first spray, the concentration in the drywell sump was 4 JC 10~ micro-1ra1115/liter ~nd the volume was approximately 332 liters. The nuober of crams of iodine transferred to the liquid in the drywell sump was 13.15 grams. The difference between the number of grams of iodine removed from the g3s phase and the number of arams of iodine transferred to the liquid phase was -8.1 grams. The resulting relative error based on the gas phase is grams removed from ~as - srams trRnsferred to liquid) error * ( - grams removed from gas (67)
JC 100% '
error * (S.OR2( 5 *-08213.15)
)
- ioo-.
- _133
- 4% * (66)
Rei;ultt> of the other ru1lS were sbihr with tnolrc iodine aripe3rin' trans-hrred t*' th!! liquid than was removed fro:n the ga'i. In fact, for many cases the error ts tnuch grt!ater.
A-40