ML20136B367
| ML20136B367 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 12/31/1976 |
| From: | ATLANTIC RESEARCH CORP. |
| To: | |
| References | |
| PB-262-180, NUDOCS 7909050311 | |
| Download: ML20136B367 (174) | |
Text
I
~
- ==
~ --
h F
3 t
U.S. OEPARTWfili 0F C0MMERCE F
~
Xatiomi Technicallnfonr.ation Sew.e l
(
l PS-262 18Q
}
i i
Hydrogen Suppression Study l
l and Testing of Ha on 1301 j
Phases I and ll i
j a
Allantic Research Corp, Alexandria, Va Applied Ph'/ sics Dept B
h l
I u
(
)
NHN j
Maritime Administration, Washington, D C f:
Dec 76 i
E J
l 4
19090503lla,
L J
w-----,-
.fB 262 180' 016003 m
I r
m HYDROGEN SUPPRESSION ST1'DY AND TESTING OF H ALON 1301 a
PHASES I AND II
[
Final Repor Edward T. \\leHale
[
Submitted by:
I L
Applied Ph>h Department ATLANTIC RESEARCH CORPOR ATION 5390 ChernLee Avenue
[
Alexandria. Virginia 22314 Submitted to:
Todd Research and Technical Dhision Post Office Box 1600 Galvesinn. Texas 77550 Attention: Str. William Shape Performed for:
%laritime Administration U.S. Department of Commerce SI A-6562 December 1976 ARC 47-5647 Contract RT-3900 6
ATLANTIC AESEAACH CC APCAATION e
3' ALEX ANC ALA.VI AGINIA. 223%
l t
I EtPRODUCID f t NATIONAL TECHNICAL l l
l INFCP)AAT'CN SERVIC2 ;
)
1 i
"ll,Wu2litc!L**
w.-
~
~
e _.. _ _
I n
l'ff NJ "[sM" #f 7h3d R epari 2.
- 3. Recipient's Acceu sen No.
sistioGaAPHic cat A d
WEET
- s. ae,or, os,e p-a.1.ac..d s..ae 0
Hydrogen Suppression Study and Testing of Halon 1301 -
Jfe; Phases I and II
- 7. Aestor(s)
- 4. Performang Organisassen Hept.
N **
7 Edward T. McHale
- f. Performaag Orgsassence home see Address IQ. Proget/ Ta ss ' loris Unit No.
Atlantic Research Corpora} ion
' C'**'* *'# 6'"' NMA 6 5 6 2 5390 Cherokee Avenue' 22314 MarAd Contract Alexandria, Virginia Todd Subcontract RT-390 )
'a
- 83. Type of Report as Period
- 12. Speemosang OrsteLaatson Name and Address Coeered Final 8
Department. of Com:serce Nov. 19 74-De c. 19 76 Maritime Administration I d' 14th and 0 Streets, N.W.
m., o me n r an.mn
- 15. Supplementary Noie s
]-
- 16. Abserscas 1.
In the event of a loss-of-coolant-accident in the reactor to be used to power nuc-i lear ships, it can be postulated that sufficiently high concentrations of H2 and 02 might develop that a resulting explosion would produce overpressures that exceed the containment design. If a LDCA occurred and the H2 concentration reached the explosive limit, an inerting gas could be discharged into the containment to suppress any possibl a explosion. This report describes a study that was conducted to evaluate the feasibilit/
{-'
of eciploying Halon 13)1 in an explosion suppression system for a maritime nuclear reac-One experimental task involved measuring the quantity of agent require'd to inert tor.
H,02, N2 mixtures over a range of conditions likely to be encountered in a contain-2 In another task, a facility was assecbled which simulated a containment vessel cient.
}'. on subscale. Several tasks were addressed to technical problems that could be en-J visioned if Halon were present in a containment vessel.
"he results of the study J-support the concept of t: sing Halon 1301 for containment protection.
- 11. Aey Soros sad Docuses: Assaysts. lle. Descripcors Hydrogen Suppression l
Halen 1301 leerting Loss-of-Coolant Protection Containment Protection Hydrogen Explosion
[
Explosion Protection
~~
1 17b Idemifier,/ Opes-Ended Terms e
o 17s. CoS ATI rield ' Group
,1
,f s7
- 19. dec urn y unsa (1 tus
- 21. N. atr'ase,
- 18. AesOsbehty 2:aienner's
'j j ve v.
P. e get )
f p>f P '" ** N T.elease t'nlimited -
- 22. rnn Available frcs NTIS if,',',i n w i = u n i.
T.5
- e t__.ee c.rn
!}tih r GRM Mn Y P L 4F RORCAD w e c osa+ u r. s a s o r' t.
- G eteis es s e e se 4 4 s e is. 7 56 Lghg gy ng,agg ( g g g, I
L
' L.
FORDIORD
{
The study covered by this report was conducted under subcontract RT-3900 f,
to Todd Research and Technical Divisica of Todd Shipyards Corporation in support of contract MA-6562 with the Department of Commerce, Maritime Administration.
I I
The period of performance extended from November 14, 1974, to December 17, 1976.
i Mr. Robert D. Cain and Mr. James E. Warwick served as the principal technical i
e representatives at Todd, and Mr. Charles P. Patterson of the Office of Advanced l
Ship Dewlopment served as the technical representative at Marad (Code 920).
Work on the project is continuing under Contract T-38169. Quality assurance i
documentation that has been collected in the course of the program is on file and available at Todd and Atlantic Rasearch Corporation.
L.
k i.
1.
t1.
I k..
n h
,o
~ O e
I.
l j
.a
.i.
i l
p 1%
=
~
1.FSA1. NOTICE This report was prepared as an account of government-sponsored work.
Neither the United States, nor the Maritime Administration, nor any person
-l acting on behalf of the Maritime Administration:
1)
Makes any warranty or representation, expressed or implied, with respect to the accuracy, completeness, or usefulness of the inforsation contained in this report, or that the use of any information, apparatus, method, or process disclosed in this report may not infringe on privately owned rights; or 2)
Assumes any liabilities with respect to the uss of, or for damages resulting from the use of any information, apparatus, method, or process disclosed in tiis n ort.
As used in the above " persons acting on ahalf of the Maritime Administration" includes any employee or contractor of the Maritime Administra-tion to the extent that such employee or contrcctor prepares, handles, or distributes, or provides access to, any infortation pursuant to t. tis erploy-ment or contract with the Maritime Administration.
i i
i l.
t I'
f i
I^
I.
I l
i T*
i l
11
__.._...s.........e.".e t
C TABLE OF CGNTENTS Page
.1. 0.
INTRODUCTIOT.
1-1 2.0 THERMOCHIMICA.L COVKTER CALCULATIONS OF EIP!ASION AND DETONATION CONDITI*55' TOR Hgg MIIIURES 2-1 1
3.0 EXPLOSION-L*.VIT v!ASUR!MENTS OF Hgg - RALON 1301 MIrtURES......................
3-1 3.1 LABORATORT-SCALI FrASURMENTS..........'....' 3-1 e
m.
i 3.2 INTERMEDIATI-SCALE AND LARCE-SCALE ICNITION TESTS 3-2 i
3.3 APPARATUS PCR.VIASGIMENT OF EIPLOSION LIMITS 3-7 L.
3.4 PROCEDURE PCI CCITCCTING EXPLOSION-LIMIT MEASCRIMENTS 3-9 4.0 LARCE-SCALE "'ESTI5C 4-1 4.1 ENVISIONED FE3LIMS 4-1 j
4.2 Dr.SCRIFFION OF TIST FACILITT...
4-2 1
4.3 RESULTS 4-5 5.0 MEASUREMENT CF A354R7 TION EFFECTIVENESS OF CHARCDAL
UPON EXPO! m "*3 ER ON 1301 5-1 5.1 ENVIRONMIS"!AI. EMINIERING & TESTING REPORT TO ATLANTIC RESEARCH C::1PCIATICN ON ABSORPTION ETTECIIVINESS OF CEARCOAL UPC5 EIPCSURE TO HALON 1301...........
5-3 i
)
6.0 RAD IO LYS I S C F E.CC T 1301.................
6-1 6.1 DISCUSSION CF EET QCESTIONS 6-1 6.1.1 Re a c t i on o f I, wit h CFg................
6-3 6.1.2 Conclusie,s fres A alysis of I, Penetions 6-7 l,
6.2 NUS CDRP01ATIC5 IIPORT TO ATY. ANTIC RESEAECH CDEPORATION ON "RADIOLTSIS OF 3MTRIFLUOROMETHANE (FXCN 1301) IN A I
COMEUSTIBLE CAS CDSTIOL SYSTIM F02 IEE HARITIME REACTOR".
6-8 7' 0 RATE OF SOL"?!I.I T OF HALON 1301 IN WATER 7-1
{
7.1 DESCRIPTION
OF P101 LIM AND PRELIMINARY MCDELI5G 7-1 7.2 EXPERIMENTAL A?PAEATUS AND KETHOD OF CHEMICAL ANALYSIS.
7-8
}
7.J EXPERlMENTAL IIST II.5ULTS 7-12 I*
7.4 DISC"SSION C7125CLTS 7-21 I
111 l
r
=1
.... -- o.;...a,,,...n m. m 1
t 4
L (TABLF OF CONTDTr5 - Continued) q O
Pago 8.0
_FRELIMINARY SYSTEM ANALYSIS 8-1 m
8.1 SYSTEM CONFIGURATION 8-1 l
<e
- 8. 2 SAMPLE CALCULATIONS OF HALON 1301 REQUIPIMENT 6-3
- 8. 3
'EFFECT OF PRESENCE OF STEAM 8-12 l
9.0' (X)NCLUSIONS...'....................
9-1 o.
i.
LIST OF TABLES 1.1 Physical Properties of Halon 1301.............
1-3 1.2 Vapor Pressure and Density of DuPont Halon 1301......
1-4 2.1 Susmary of Calculations of Explosion and Detonation Pres-sures and Temperature for H -Oxidant Mixtures..
2-2 2
3.1 Sungsary of Explosion-Limit Results for Tests Conducted in 5.6 Cubic Foot Sphere (12'-1 through 13'-4) 3-4
- 3. 2 Summary of Explosion-Limit Results for 'l--te conducted in 5.6 Cubic Foot Sphere (13'-5 through 14'-5) 3-5 3.3 Summary of Explosion-Limit Results for Tests Conducted i(.
in 5.6 Cubic Foot Sphere (14'-6 through 15'-4) 3-6 5.1 Performance Recuirements 5-3 1
7.1 Summary of Test Conditions for Solubility Rate Study 7 12 i
~
- 7. 2 Experimenta1 F-Factors in Solubility Rate Study 7-2 J i
- 7. 3 Summary of Average F-Factors for Various Conditions in
~*
Solubility Rate Study...................
7-24 e
T 1
)
iv 1
1
~
~ ~
~
i l
.a LIST OF FICUPI.S l
m.
Page 2.1 Calculated Adiabatic Constant Volt:ne fxplosion and l'
Chaptnan-Jouget Detonacion Pressurea and Temperatures For H -Air.
Initial P = 14. 7 psia T = 29 8'K......
2-3 2
3.1 Schematic of Apparatus' For B Explosion Limit Measurements.
3-12 2
3.2 Explosion Limits, H -Air-1301 2
(accompanied by tabulated data) 3-l's j
3.3 Explosion Limits, H -Air-1331 2
(accompanied by tabulated data).........
'3-15 3.4 Explosion Limits, H -N /0, 2/1-1301
~
3 2 2 (accompanied by tabulated data) 3-17 j
- 3. 5.
Explosion Limits. H -N /0, 2/1-1301 l
2 2 2 (accompanied by tabulated deta) 3-20 I
4 3.6 Explosion Limits, H -N /0,1/1-1301 2 2 2 (accompanied by tabulated data).......
3-22 1
3.7 Explosion Limits. H -H /0,1/1-1301 l
2 2 7 j
'(accompanied by tabulated data).......
3-25
)
f
- 3. 8 Explosion Limits, H ~N /03, 1/1-1301
(
5 2 2 (accompanied by tabulated data) 3-27 f
n.
3.9 Explosion Limits, H "N /0 ' I/1~1301 2 2 2 j,
(accontpanied by tabulated data)...
3-29 3.10 Explosion Limits, H -N /0,1/1-1301 2 y 2 L
(accompanied by tabulated data) 3-31 3.11 Halon 1301 Peak Percentage For Suppression of Erplosions
"~
of Hydrogen and 0 /N Mixturn (7 *F; 1 atm).
3-33 2 2 i
3.12 Plot Showing Measured Pressure Increase Wen Mixtures of H and 1:1/N :0 Containing Halon 1301 Were Ignited y
2 2 t..
in a 5.6 cu f t Spherical Vessel..............
3-34 e
r-4.1 Schematic Diagram of Test Facility L* sed to Conduct Large-l Scale Halon 1301 Vaporizatien and Ignition Tests 4-3 4.2 Plot Showing Pressure Increase as a Function of Time in Large Volume Tank for Test in ~.'hich Liquified Halon 1301 l
Was Inj ected a t Rate of 20.0 lb/cin............
4-7 i
v J
l
a 6
(LIST OF-FIGURES - Continued) 8 Page lj 4.3 Large-Scale Test 3.
Plot Showing Pressure Increaue k!
as a Function of Time in Large Volume Chanber Fer Test in Which Liquified llaJon 1301 Was Injected at Rate of 13 lb/ min... !.....................
4-8 4.4 Large-Scale Test 4.
Plot Showing Pressure Increase as a Function of Time in Large Volume Chamber For Test in Which Liquified Halon 1301 Wae Injected at Rate of u
33 lb/ min.........................
4-9 L*
7.1 Computer Printed Concentration Profile for Dif fusion-Limited Hodal of Halon 1301 Dissolution in H O (N=10~0)..
7-4 2
722 Computer Printed Concentration Profile For Diffusion-Limited Model of Halon 1301 Dissolut % in H o (H-10~0)..
7-5 y
7.3 Plot of the Quantity N Versus Area A of Concentration Profile Curves in Solubility Rate Model..........
7-6 j
7.4 Sphematic Diagram of Apparatus Used For Study of 7-9 Solubility Rate of Halon 1301 in Water 7.5 Plot Showing Ef fect On H O Flow Rate On Solubility 2
Rate of Halon 1301 in 110 (D-value For Theoretical Lines 2
-5 is 10 cm'/sec)...................
7-15 7.6 Rate of Solubility of Halon 1301 in 110 During 2
7-16 Quiescent Test......................
1+
- 7. 7 Plot Showing Effect of Halon 1301 Partial Pressure ca 4
7-17 Solubility Rate and on Equilibrium Concentration
- 7. 8 Plot Showing Fquilibrium Concentratic.n of flaien 1301 in H O at 120*F and 0.5 and 1.0 atm Pressure......
7-18 2
7.9 Effect of Splseh Suppressor on Solubility Race of Halon 7-19 1J01 ir !! 0 in Test Apparatus....
2 7.10 Comparison of Pata ? rom Test 6 (3 gpm H 0; 76*F; 2
i.
1.0 atm 1301) With Theory...
7-20 l
i.
4 s.
2.
7 vi
f
._J I
J (LIST OF FICURES - Continued) 1
{
Page
(
u.
j 8.1 Pos t-LOCA Combustible Gas Cont rol System.
B-2 t
i i
8.3 Pressure Buildup in Containment Following LOCA
{
i i
For ARC Sampla Case B-6 e
S.4 Halon 1301 Requirement For Containment Inerting For ARC Sample Case, Based on Peak Percents 3e B-7 i L}
- 8. 5 Partial Pressure of H2 ""d T*E"1 Pressure in Containment Following LOCA if No Halon 1301 is Injected B-9
]
8.6 Halon 1301 Requirement For Containment Inerting For ARC Sample Case Based on Maintaining H Concentration 2
I f
B e l ow 4. 0 %......................... 8-10 8.7 Percent H in Containment Following IECA With No 1301 2
1 Injection and With Injection at Constant 12 lb/ min..... G-11 l
' l
) '
8.8 Logic Diagram Illustrating Procedure For Calculating l
Halon 1301 Requirement for a Containment Vessel i
l Following a LOCA.
B-13 a
f 8.9 Preliminary Estimates by Babcock and Wilcox of Blowdown Conditions in Containment B-14
- i.
I i
1 I
e 4
1 I
t*
I em vfl I
~
- ~~~~
A Ll f
I
SUMMARY
A loss-of-coolant-accident in a water-cooled nuclear reactor can result in the accumulation of hydrogen and oxygen gases in the containment vessel. In the case of the reactor to bo used to power nuclear ships (Babcock and Wilcox CNSC),
using the conservative C values of USNRC Re2ulatory Guide 1.7, it can be postulated L
that sufficiently high concentrations of H end 02 ight develop that a resulting 2
explosion would produce overpressures that exceed t';e containment design. One means f
of protec51on against such a hazard ia incorporation of an inerting system into the design. If a LOCA occur e.1 and the H concentrat10 ! rer.ched the expliesive limit, an 2
icerting gas would be discharged into the containment to suppress any pcssible ex-i plosion.
Halon 1301 (0F Br) is the most sMle inerting agent for the contain-3
~
ment application. h agent would be stored in the '11guf fied statr and injected j
only if needed. The Halon system offers operational advantages of effectiv.: ness.
reliability, reistive economy, storage convenience, mechanical simplicity, few noving parts, minimal power requirements, ease of priodic testing, etc. Bis report desefibes a study that was conducted to evaluate the feasibility of em-playi=g Ealon 1301 in an explosion suppression system for a maritime nuclear
{_
reactor.
I The first experimental task of the program involved measuring the quan-l, tic 7 of agent required to inert H, 0, H mixtures of a range of comp sitions and 2
2 2
at conditions of pressure and temperature likely to be encountered in a contain-r-
cent. Explcsion suppression tests were conducted on laboratory, intermediate,
and large-scale. In another task of the program, a facility was assembled which j
simulated a containment vessel on subscale. Agent injection and vaporization tr.sts vere performed under conditions expected to be present in a containment.
Several tasks were addressed to technical problems that could be en-visicced if Halen were present in a containment vessel. In one of these, charcoal used in the filter system of the containment was investigated for its effectiveness in alsorbing radioactive iodine co= pounds in the presence of Halon 1301. Another question concerned the stability of gasec us 1301 in the radiation field of the cont 41:sent over a 67-day period if the sgent had to be injected following a LOCA.
It was shewn that neither of these potential proble=s was significant.
1<
_ ~.
o,
~
,-.ar e scrp e m :,w vw-a w.~.>a m
==%
i a
rw i,
j l
It was found, however, that ene envisioned problem could be potentially aerious. This concerns the dissolution of 1301 11 emergency cooling water with Even its subsequent deco = position as the water passes through the reactor core.
though the solubility of Halon in we.*i r is low, if the dissolveo quantity decom-4 posed as it passed through the core and was replenished as the water. ras reexposed to the Halon, it is possible to estimate a large loss of agent over a 67-day pe riod. There are two r.spects to the problem - the rate of solubility of 2 301 in H 0, and secondly the radiolytic dec.omposition of 1301 in solutiot..
In the 2
pre.sent program th. rate of solubility of Halon in water was studtud rad found to be appreciable.
'Ihis, therefore, requires a study of radiolysis of 1301 which la planned for the next phase of the program.
The report also contains an anal / sis of the Halon 1301 system desiga for 1
applicat' ion in the CNSG. Io a sample computation it is calculated that a 70,000 f t' 3
containment of 35,000 ft ullage would require of the order of 20,000 po' rads of l
agent (23 psia) to inert against the worst hazard that could occur base 1 on CSNRC Regulatory Guide 1.7.
~41th the exception of the ensettled uater decomp 3sitio.2 ques-
~
tion, the results of the study very favorably indicate that a Halon 1301 sunression system is ideally suited for a nuclear containment application.
\\
1 i
i i
l.
rl.
t g
i i..
=
7
{
i f
1.0 INTRODitCTION i
In the event of a loss-of-coolant-accident (LOCA) in a water-cooled nuclear reactor, hydrogen gas accompanied by oxygen can develop in the contain-ment vessel. The principal sources of the gases are radiolysis of the emergency cooling water and the reaction of H o with zirconium cladding of the reactor.
y In a maritime reactor (the Babcock and Wilcox Consolidated Nuclear Steam Cenera-
. [
tor (CNSC)), to be used for propulsion of nuclear powered ships, the containment 3
volume is of the order of 70,000 f t, of which about 35,000 ft is ullage. H 2
{
and 0 concentrations thet are well into the explosive range can accumulata in 2
such a containment. Ignition of these gas mixtures could lead to overpressures l
that exceed the capacity of the containment design.
I, One means of procuction against such a hazard following a 1.0CA is to e
discharge an inerting gas into the containment. Once mixed with the flammable gases, the inertant would suppress explosion of hydrogen-oxygen mixtures of any composition. An agent known as Halon 1301, whose chemical fortsula is CF Br, 3
was selected as the inertant. The program described in this report addressed many technien1 questions associated with the use of a Halon 1301 explosion l
suppression, system for application in a CNSG containment vessel.
The suppression system would ccanist of a predetermined quantity of Halon 1301 stored as a liquified gas in several storage vessela near the con-tainment. In the event of a LOCA, if the hydrogen concentration reached the l
lower explosion limit of 4 perc-nt by volume, the 1301 would be discharged either automatically or manu:11y into the containment through a piping and valve mystem. Many operational advantages are associated with the sicplicity of a 1301 system. For example, there are few moving parts, minimal power requirements, and once activated the eystem requires little further attention.
e Typically, about 20,000 pounds of the agent would be required to inert
}
3 35,000 ft free volume containing the concentrations of 11, 0 and N that one j
2 2
2 computes based on USNRC Regulatory Guide 1.7. This cuantity of Ha.lon would produce
{
a pressure in the containn.ent of just over 20 psia. One of the nain tasks of
(
the program was to determine the exact amount of 1201 required for a range of conditions and H /02 " "" * " " #' E I "' '
2 e-i 1-1 g
L.
In principle, it is possible to use gases other t! an Halon 1301 for L;
injecting such as nitrogen or carbon dioarde. la practice, however, the storage requirements and the pressures that other gases vould produce in the coctainment I
j vouLd be excessive. For example, assuming that the total required weight of N 2 or CO v uld be same as that of 1301 (it vould actually be somewhat higher), the 2
pressures produced in the containment would be about 130 psia for N and 80 psia f
g for CO. Storsso vessels would be required to withstand the high pressures that 2
l these apants would produce 2n the gaseous state since the critical temperatures of I'
both are lov (88'F for kI)2 and -233*F for N ).
Even storage at severe.1 thousand 2
poi pressure vould require many large heavy-valled vessels. Halons other than 1301, L
sue'h as Halon 2402 or Halon 1211. are also t.heoretical possibilities. However, on the basis of physical properties, stability. toxicology etc.. Halon 1301 emerges as superior.
Halon 1311 is well suited for the present CHSG containment vesaal appli-cation because of its physical and chemical properties and because of its favorable toxicological characteris tics. To illustrate this, some of its more important
)
properties and some results of toxicity studies can be cited. Most of these data can be found in the Halon 1301 product bulletin (No. B-29C) of DuPont, (who is the leading manufacturer of the agent and has performed extensive testing of the material). or in the NFPA Standard No.12A. cited later in this report.
A number of relevant physical properties are collected in Table 1.1.
i It can be seen from the tabulation that Ealon 1301 can be readily stored as a liquified gas under the conditions in the reactor compartrent. Fo r exa=plo,
at 120'F the vapor pressure is approximate.ly 400 psia. Its critical te=perature is sufficiently high so as to pose no prcbleng, and the liquid density even at the critical point is comparabl.: to that of E 0 (%25 percent less). In Table 1.2 2
are listed vapor pressure and density data for 1301 as a function of temperature.
The chemical properties of Halon 1301 of interest concern its compati-bility with materials of construction, its reactivity and thernal stability.
Fluorinated hydrocarbons are among tne most inert and stable chemicals cova.
As a general rule, except for reactivity with certain active eetals such as alkali metals,1301 can be considered cecpatible and non-reactive with almost all mate-rials, and storable indefinitely even at elevated temperature. For eranple, DuPont conducted tests wherein liquid 1301 was sealed into glass tubes containing
[
test strips of common metals fer a 44-month period at temperatures of 120*F and 250*F.
As a result of these tests they report that the following metals are suitable for ese with 1301:
'b
}
1-2 i
I a
1 Tabla 1. A.
Physical Properties of Halon 1301 Chemical Formula CF Br 3
Holecular Vaight 149 0
[,
Boiling Point
-72*F Freezing Point
-270*F Critical Temperature 152.6'F i
j Critical Pressure 575' psia Critical Density 4 6..', lbs/cu ft t
,,j Heat of Vaporization 35.5 BTU /lb @ 70*F Surfaca Tension of Liquid 9.7 dynes /cm @ O'F L}
4.5 drnes/cm 6 70'F Density: Liquid 82.2 lbs/cu ft 0 120*F 98.0 lbs/cu ft 0 70*F Saturated vapor 16.2 lbs/cu ft @ 120'F
- 7. 4 lbs/cu ft 0 70*F Heat Capacity 0.257 BTU /lb 'F G 120*F 1
0.205 BTU /lb 'T @ 70'F 0.176 BTU /lb *F @ O'F Vapor Pressura 400.4 psia @ 120*F 213.7 psia 0 70*F 71.2 psia 0 0*F i
Thermal Conductivity 0.025 BTU f t/hr 'F f t2 @ 70'F i
Solubility in H O 100 ppm at 15 psia and 120'F 2
i 1-3 S
l l'
1
[i l
l Table 1.2.
Vapor Pressure and Density of DuPont Halon 1301 8.
Temperature Vapor Pressure Density (ib/eu f t) e.
(*F)
(psig)
Liqui d Sat'd vapor vapor 01 atm
{
l 0
56.5 112.5 2.414 0.453 l
5 63.2 111.6 2.635 0.448 10 70.3 110.6 2.872 0.443 15 77.9 109.7 3.125 0.438
'l 20 86.1 108.7 3.397 0.433 i
j 25 94.7 107.7 3.688 0.429 30 103.9 106.7 4.000 0.424 35 113.6 105.7 4.334 0.420
~~
40 123.9 104.7 4.691 0.415 45 1 34. 8 103.6 5.073 0.411 r.
50 146.3 102.5 5.482 0.407 g
55 15 8.5 101.4 5.921 0.403 60 171.3 100.2 6.391 0.399 65 184.8 99.0 6.896 0.394 70 199.0 97.8 7.439 0.391 75 214.0 96.5 8.022 0.387 80 -
229.7 95.2 8.651 0.383 85 246.1 93.9 9.331 0.379 90 263.4 92.5 10.07 0.376 95 281.6 91.0 10.86 0.372 100 300. 6 89.4 11.73 0.369 t'
[ ~-
105 320.4 87.8 12.68 0.365 11 0 341.2 86.1 13.73 0.362 115 363.0 84.2 14.89 0.359 120 385.7 82.2 16.18 0.356 125 409.4 80.0 17.64 0.353 130 434.2 77.6 19.31 0.349 135 460.1 74.9 21.26 0.346 140 487.1 71.6 23.61 0.343 145 515.2 6's. 4 26.50 0.341 150 544.5 60.7 30.89 0.338 152.6 5'00.2 46.5 46.5 I
(critical) e 6<
t-1-.,
_ _.. ~
Stainless Steel 302 Aluminum 6061 I
Steel 1020 C2 Yellow Erass Stainless Steel 321 Magnesium AZ-91C l
Aluminum 1100 Commercial Titanium Aluminum 2024 Titanius A110 AT (Hetals not included in this list, such as copper, are not known to be incompat-f!
ible, but are omitted becadse they were not tes ted.) Thus, in the event of a l
IDCA, Halon 1301 could remain in the presence of hot metal for extended periods.
The toxicity of Halon 1301 has been a subject of extensive investigation 4
for mary years and involved both animal studies and husen exposure (see DuPont t
product bulletin No. S-35A). There are two considerations in this natters one concerns the toxicity of the neat agent; and the other concerns the toxicity of i
decomposition products. The latter include hydrogen fluoride which is highly 1
toxic (and corrosive), but is not a consideration for the present application, since HF could only arise a.f ter discharge of 1301 following an LOCA. Pe rsonnel would not come into contact with any HF produced in the containment vessel in such an event (and the relatively r, mall amount of ET expected would not represent a corrosion problem). Hence, only the toxicity of the neat agent is of interest for the CNSC containment vessel.
Ordinarilj, personnel would not como into contact with neat 1331 because it would be stored in pressurized vessels and discharged only in the event of a LOCA. If it were inndvertently discharged when there was no need for the agent, i
it is conceivabic that hu ans would be exposed to 1301 vapor. Hence, its toxicn-logical characteristics should be mentioned.
I-Halon 1301 is not censidered a toxic gas. Underwriters' Laboratories classified it in their least toxic Group 6 ("Oas or vapors which in concentra-t tionn up to at least 20 percent by volume for durations of exposure of the order I
of 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> do not appear to produce injury.") Carbcu dioxide, for m ple, is rated more toxic and placed in Group 3.
The classification is based en animal tests. There is no doubt that Halon 1301 is the least toxic agent that could be j
chosen for the containment vessel application, which would also meet all the requirements of the totalpr:.blem. However, in spite of the low toxicity of 1301 personnel would have te be advised to leave the i=ediate vicinity in the event of an accidental discharge in a confined space and to await return until af ter the area had been ventilated. In ger.eral, it can be assured that 1301 toxicity would present no hindrance to choosing this type of inerting system.
(
1-5
(~
The program that was co'nducted to evaluate the technical feasibility of empicying a Halon 1301 system for containment protection is reported in the following sections. Section 2 summarizes results of theoretical calculations of explosion and detonation pressures and temperatures that would be produced by various H, 0, N mixtures. Secti n 3 reports experimental reshits of explosion-3 2
2 limit mapping of mixtures of H, 0, N and Halon 1301. The amount of 1301 re-2 2
2 quired to inert various gas mixtures under a range of conditions was measured in testa on laboratory. intermediate, and large-scale.
l*
L.
In Sectior.s 4 through 7, specific problems were investigatec that were envisioned might arise in the application of a 1301 system. In Section 4 a des-cription is given of essentially a miniature containment system which was assembled and in which simulated post-LOCA tests were performed. Section 5 reports results of testa to determine whether containment filter systems are affected by the pres-ence of 1301. The question of the stability of 1301 in the radiation field of g+
the containment following a LOCA is considered in Section 6.
Lastly, Section 7
- I reports on a study 'of the rate at which 1301 dissolves in water. Section 8 com-prises an analysis of a typical application of 1301 to a containment. Sample calculat. ions are given of how to ' compute the required quantity of agent and the rate at which injection cust occur to insure protection against explosion.
t 4
L I[
1-6 8<
d
--y
.x
--n---w,%
~.
O I
2.0 THET:MOCHDfICU, COMPUTER CALCI 1ATIONS OF EXPLOSION AND DETONATION CONDITIONS FOR H; - Ng MIITURES Theoretical emplosioco and detonation pressures and temperatures that would result from the ignition of H -exidant mixtures of various compositions 2
have been calculated and are preaanted in Table 2.1.
Adiabatic constant volume conditions were taken for the explosion processes, and the detonation values represent ideal Qiapasn-Jouget procesaas. An initial temperature of 25*C was used, and thermochemical data for the calculations were obtained from the JANNAF source '(JANNAF Thermochemical Tables NSRDS-NBS37, Government Printing Office.
Catalog No. C 13.48:37).
I l
There are an infinite number of mixtures of H - xidant that could be 2
considered, and calculations were made for 20 selected cases listed in Table 2.1.
The first group of cases (1-8) refer to air at an initial pressure of 14.7 psia-
)
the secend and third. groups (9-14) refer to N /0 = 2.0 at 14.7 and 29.4 psia; 2 2 the fifth and sixth groups refer to E /0 = 1.0 at 14.7 and 29.4 psia. The re-2 2 suits for H -Air a tures are plotted in Figure 2.1.
2 Mixture cospositions were -chosen somewhat arbitrarily: Cases 1 and 8
)
represent *the lower and upper explosion limits of H in air; Case 4 represents 2
approximately the peak percentage of H ; case 5 is the stoichiometric mixture 2
i of H in air; Case 6 is the stoichiometric mixture saturated with H O vapor.
y 2
The detonation limits of H in air are 18 (1 wer) and 59 (upper) volume percent, 2
hence no detonation results are rhown for mixtures outside these limits, even though it is possible to theoretically compute such results. It is noted that the reflected shock wave associated with a detonation would approximately double the detonation pressures given in Teble 2.1 The oxygen-enriched mixtures (Cases 9-20) that were chosen represent approx 1=ately peak percentages of E2 (Cases 9,12,15,18); peak percentages saturated with H O vapor (Cases 10,'13, 16, 19); and stoichiometric mixtures, 2
H /02 = 2:1 (Cases 11,14,17, 20).
2 The calculations represent theoretical combustion conditions, but in any real sitaation the behavior of amplosive hydrogen mixtures 'depc-ts widely from the ideal case. At low hydrogen concentrations (less than about 6 percent) a special type of ec bustion occurs, known as cellular flame propagation, i
2-1 O<
t
- - ~,. _.
- 1
!~!
M M
- i. i
...A
!~"~.1 aC
- i..i
.. 4 L __4 L.._.
C_~2 L.
E '~
~
I -
T-'
8 Table 2.1 Summary of Calculations of Explosion and Dotonation Pressures and Temperatures for 11 - Oxidant Mixtures.
2 Initial Explosion Explosion Detonation Detonation
' Pressure Temperature Pressure Temperature Volumetric Percent Ratio Pressure Case
- Jy_,
4_
J-S 0, Li g _
_(psia)
(psla)
(*K)
(esta)
(K) 2 2
- 3.8 14.7 36 745
'l 4
20.1 75.9 3.8 14.7 45 950 2
6 19.6 74.4 3.8 14.7 63 1339 3
10 18.8 71.2 4
15 17.8 67.2 3.8 14.7 82 1790 5
29.5 14.7 55.8 3.8 14.7 118 2748 2?.9 2944 6
26.3 13.1 49.6 11 3.8 14.7 10 9 2519 212 2711 3.8 14.7 104 2351 202 2597 7
50 10.5 39.5 8
75 5.2 19.8 3.8 14.7 67 1436 9
15 28.3 56.7 2.0 14.7 81 1778 160 1978 10 13.4 25.2 50.4 11 2.0 14.7 74 1600 141 1782 2.0 14.7 129 3079 251 3258 11 40 20 40 12 15 28.3 56.7 2.0 29.4 162 1778 311 1979 h'
13 13.4 25.2 50.4 11 2.0 29.4 147 1600 281 1782 CD A
14 40 20 40 2.0 29.4 261 3148 511 3339 1.0 14.7 B0 1763 154 1960 15 15 42.5 42.5 16 13.4 37.8 37.8 11 1.0 14.7 73 1589 140 1768 1.0 14.7 135 3277 263 3452 17 50 25 25 1.0 29.4 161 1763 309 1961 18 15 42.5 42.5 i
19 13.4 37.8 37.6 11 1.0 29.4 146 1589 279 1769 l
20 50 25 25 1.0 29.4 275 3369 537 3556 i
e
=
l 7-.-
ri r - --
r --
r r.--
r-L--a l
r 300 3000 Detonation T
~~~,'s s-
/
N
\\
250 2500
[',',
' w N
N s
s '
Explosion T s
200 Detonation P 2000
/
y
,' /
\\,,
E.
s w
s z
ad P
1500 F E 150 s'
N 3
m 1
4 6"
l m
to h8 s
a.
, ~~ ~~-
,,'Emplosion P 100 t
a.
,~ '
1000
/
o
~~~~ ~
~~
f 50
/--
500 s'
0 0
0 10 20 30 40 50 60 70 60 VOLUMETRIC PERCENT H2 IN AIR l
Figure 2.l. Calci.,.ed Adiabatic Constant Volume Emplosion and Chapman-Jouget Detonation Pressures and Temperatures for H Air. Initlet P = I4.7 psla, T = 298*K.
2
(
i 4
4
l 1
1 l i'.
l 4
sa
! as,
f; wherein only a fraction of the H is cor.sumed and the theoretical pressures and 2
temperatures are not nearly achieved. With higher concentrations of H i"
- 2 large volume system, the rate of explosion propagation would accelerate af ter initiation. Even though the H concentration was below 18 percent (the lower 2
l detonation limit) the rapid reaction would give rise to shock waves which, while 4,
!L not producing tne detonation pressures, might neverthelese cause pressures in excess of the thaoretically computed explosion values. Thus, once the H con-2 centration exceeds approximately 6 percent, it becomes critical that the con-tainment be inerted.
fu l (L f
6 i
n! l I
- i.,
i...
I 1..
J i
I e
i I
e i
12<
2-a
=
i 3.0 EXPLOSION-LIMIT MEASURD'ETTS OF Hgg-EAIAN 1301 MTIWRES i,
l.,
3.1 1.ABORATORY-SCALE MF).6CEEXINTS The purpose of this task of the program was to measure the quantities of Balon 1301 (CF Br) required to ir.a rt d -0 ~v
t**"#** *# **#I""* ***E *i~
3 2 2 2 tions and at several initial tapsperatures and pressures. These results would fj then be used to specify th's amount of 1301.raqu.f red.for any given set of condi-tions in a nuclear containment application, A standardized procedura was employed which has been developed by the U. 5. Buesau of Mines and which is L.
accepted as a method to determine explosion limits for hisardous industrial proces ses.
The results from the standardised procedure were verified in a isrser-scale apparatus, and, lastly, ignition tests were performed in a 1200 cubic foot tank for final verification.
Essentially the Bureau of Minas procedure consists of preparing H N mixtures in a 2-inch by 48-inch vertically-motored tube and determining 2 2 2 o.
d th amount of Halon 1301 that must be added to render the misture non-fla:nmable when. an igniter is activated in the tube. Complete descriptions are given in j
folleuing sections of both the apparatur and the test procedures, so they need not j
be repeated *ite re.
A schematic diagram of the 1doratory apparatus I'm shown in j
~
f; yigure 3.1.
The descriptions also comprise the Quality Assuraece documentation for the test procedures used in the explosive-lit:it ceasurements. The conditions under which.he tests were performed are eu:anarized below (tnulations and plote of the ni data are collected at the end of Section 3).
Oxidant Initial Initia.1
~
j (Hominal Temperature Pressure Ignitica Htatidity Figure E g Ratto)
(*F)
(At=)
Source Conditions..
Referenen j
Air 70 1
Squib Dry cases 3.2 Air 70 1
Spark Dry Cases 3.3
~
2:1 120 1
Squib Dry cases 3.4 2:1 70 1
Spark Dry cases 3.5 7-1:2 70 1
Squib Dry Cases 3.6 i
e 1:1 70 1
Spark Dry casee 3.7 1:1 120 1
Squib Dry Casee 3.8 l
1:1 120 1
Squib H 0 Saturated 3.9 1
1:1 120 3
Squib Dry cases 3.10 f
31 13<
o l
i The exact N /0 ati 8 corresponding to the 2:1 and 1:1 nominal values 2 2 listed above were 1.97:1 and 1.0811, respectively. The program called for con-ducting tests only with squib igniters, however, in several cases spark ignition was useo also for comparison. The numbers listed in the last column designated figure reference refer to the accompanying plots of th*: explosion results. Data of all the tests are collected in the tables that accoctpany the figures. These
.1 data cemprise the composition of each mixture tested and the test reJults.
Figures 3.2 through 3.10 are referred to as explosion diagrame in tri-b angular coordinates. The dashed lines on the diagrams are the lines along which the H /0 rati e are lil and 2:1. The latter is the stoichiometric ratio; the 7
2 2
.L former is close to the composition where the " peak percentages" of Halon 1301 occur. The peak percentage represents the amount of 1301 that must be present r-
{
to prevent the explosion of mixtures of any composition for the particular H - xidant system. A summary plot of peak percentage as a function of H /0 2
2 2 i'
ratio is presented in Figure 3.11.
The lines refer to 70*F and 1 atm. No
~'
measurable temperature influence was found for the peak percentages within the i
range tested (coorpare Figures 3.6 and 3.8).
However, a small increase with in-creasing pressure and a decrease with H O saturation were measured, and these 2
effects are shown in Figure 3.11.
Figure 3.11 represents the key to specifying the amount of dalen 1301 that would be required to inert a containment if H gas (secompanied by 0 ) ****
y 2
produced following a LOCA. In order to use Figure 3.11, all eceditione asso-ciated with a LOCA must he known in advance, such as volume of containment dry and wet wella, maximum amount of H; and 02 that can be produced, initial air pressure before LOCA, etc. From this informaticu a final N /0 reti can be 2 2 computed. Den from Figure 3.11, the valuestric percentage of 1301 required
~~
for inerting can be read. A detailed illustrative example is given in a later section of this report, 3.2 INTERMEDIATE-SCALE AND LARGE-SCALE IGNITION TESTS a
The results obtained in the 48-inch tube were verified using a 5.6 cubic foot explosion vessel. The vessel was essentially spherical in shape and
~
its size nasured that no " wall effects" would be present. The test setup was very similar in arrangement to t. hat shewn in Figure 3.1.
The difference was 3-2 14<
I
~ ~..
'I 6
l 4
i L.
I that the explosion tube was replaced with the 5.6 cubic foot fiber glass sphere.
The sphere had a working pressure of 13 psi and a test pressure of 300 pai, al-though all tests were performed at an initial pressure of one atmosphere. In addition the squib igniter was positioned in the center of the sphere, and the mixing pacp was replaced with a circulating fan for *.1xing of the gases. The circulating fan had 3-inch bisdes and was run by a small AC motor. The fan shaft L.
was long enough to extend into the center of the sphere.
The mixing procedure was the s'ame as with the tube apparatus. Tests were not conducted with non-inerted mixtures, so no explosions propagated.
However, as the quantity of Halo.,1301 was reduced, some weak reaction would
.tj initially occur in the vicinity of the ignition source and then would immediately self-quench. Thi9 would produce a small pressure rise which was monitored with a pressure transducer that was mounted in the wall of the spherical vessel.
)
i Tests were conducted with the three oxidants - air, 2:1/N :0, and 2 3 1:1/N :0 - near the " peak percentage" composition of each. These correspond 2 2 to the cases shown in Figures 3.2, 3.4 and 3.6 (70*F rather than 120*F in
{
Figure 3.4).
The peak percentages of Halon 1301 for these cases were approxi-mately 24, #4 and 58, respectively. Mixtures near these co= positions were pre-pared, igniters fired, and the pressure increase measured. The results are I
collected in Tr.bles 3.1, 3.2 and 3.3.
The pressure rise observed will depend on the size of the test vessel. These same r.ixtures would produce no measurable
{,
pressure rise in a much larger vessel. In most of the tests the amount of hy-ratio was near unity. The results of drogen was fairly constant and the H;/02 i
Table 3.3 are plotted for illustration purposes in Figure 3.12.
This plot shows how pressure increase develops as the quantity of 1301 is reduced in a U /N /0 mixture. Further reduction in 1301 below 55 percent would eventually 2 2 2 lead to a fully developed explosion producing an overpressure of the order of 100 psi or greater.
The results of this intemediate-scale testing in a 5.6 cubic foot vessel' essentially verified the smaller-scale data and the explosion-limit diagrams of Figures 3.2 through 3.11.
The results shown in Figure 3.11 can be used for design of explosion suppression systa=s. Approximately a ten per-cent excess of Halon 1301 over that required by Figure 3.11 would represent a desirable safety =argin. The final verification of the explosion-limit results was made by igniting inerted nixtures in large-scale tests using a
'3 1.
i 154 1
~ - -
o l
t t
o
+
.,J t
Table 3.1.
Summary of Explocion-Limit Results for Tests Conducted in 5.6 Cubie Foot Sphere Pressure Rise Test
- I 1301
% Air
- H, (PSIA)
-12'-1 27,.0
- 60.2 12.8 0.2 12'-2 24.0 62.0 14.0
- 1. 8
~j..
,a 12'-3 25.1 61.0 13.9 1.2
>i 13'-1 26.1 60.1 13.7 0.4 l
l 13'-2 24.3 59.7 16.1 1.8 i
13'-3 24.2 63.9 11.9 0.15 l'
. 13'-4 23.0 62.0 15.0 2.6 j
i I.
I P = 1 ata T = 20*C c
1 I
e 1.
Tr.e test numbers listad above in this and all tables of See;1on 3 refer to lahorntory notebook entries. The apostrophe designattor.s f.
rr/rasent, entries in a second notebook.
'l 4
4 6
I
- k 16<
34 r
m
. -. - ~
1 a
1 1.
Tabic 3.2.
Summary of Explosi.W,ait Rasults for Testa condt.cted in 5.6 Cd ic Fat Sphare Pressure Rise Test
! 1301 2 2/1 Ng Q
(PSIA) 13'-5 44.8 42.1 13.1 0.4
,)
14'-1 44.1 42 1 13.9 0.8 h
14'-2 42.8 43.0 14.2 1.1
)
a 14'-3 42.8 45.1 12.1 0.55 j
T' 14'-4.
43.0 41.1 15 9 15 l
14'-5 46.0 40.0 14.0 0.2 j
i l i
4 L
i i
P = 1 ata
! =
3* C i
c 1
1 l
j t.
i i
i t
1 t
e is 4~
Ae<
y,3 t
g
I l
1 I
l l
. Table 3. 3.
Summary of Explosion-1.init Rasults for
~
Testa Conducted in 5.6 Cubie Foot Sphere Pressure Rise j
_ PSIA)
(
L Test 2 1301 21/1 Ng Q
14'-6 59.0,
26.0 15.0 0.1 a
15'-1 56.8 28.0 15.2 0.55 i
~
,]
15'-2 56.0 27.0 17.0 0.75 E, i, 15'-3 54.9 29.9 15.2 2.4
~
t 15'-4 55.8 30.0 14.2 0.9 I
P = 1 atm T = 00'C c
g 1
s l
1.
18<
i 3-6 i
B a
,n
-.-n, e<
r,-s
I i)'
Ir L
1200 cubic foot tank (see Figure 4.1).
The teak was the same vessel employed I
in the large-scale tests, which are described in the next section of this report.
L
'in this section a description of the large-scale ignition tests will be given.
The general procedure followed in preparing gas mixtures in the 1200 cubic foot vessel was similar to that used in both smaller-scale stucies. Air
}
vas pumped out of the vessel to a predetermined pressure level. If roquired, mixture. Following tt is.
oxygen was then admitted to prepare a 2:1/N :02 2
Halon 1301 was admitted to a predetermined pressura, and finally hydregen ges was admitted to bring the final pressure exactly to s=bient. The gases were mixed for a few minutes by a 1000 CFM circulating fan located in the vessel.
g.
[
Three mixtures of the following composition were tested:
Air 58 %
Halon 1301 28 %
H 14 Z 2
h,i t.:
i 2:1/N :0 40 %
y 2 Halon 1301 46 %
H 14 2
- I.
s.
Air 59 %
U Raion 1301 27 %
lU H
14 Z 2
l
)
In each of these mixtures, two squib igniters, suspended in the tank, lq vere fired. During the igniter firings, pressure was monitored with a trans-
}d ducer. No measurable pressure increases were observed. Motion pictures taken f
during the tests showed that the igniters fired and no other reaction occurred.
u
-i 3.3 AFPARATUS MR }EASUREMENT OF EXPLOSION LIMITS A schematic diagram of the apparatus used for the measurement of explo-sion limits of H ~N /0 -Halen 1301 mixtures is shown in Figure 3.1.,
The design 2 2 2 is adapted from that of the standardized apparatus developed by the U.S. Bureau 37 19 <.
- -++we
- W Me emme n w
m,m.. m
-ee9
I of Mines for flame. ability-limit measurements *. Following is a description of the
{
main features of the appsratus that was constructed under subcontract RT-3900, b
"A Hydrogen Suppression Study and Large-Scale Testing of Halon 1301," TodJ Research and Technical Divisios, Todd Shipyards Corporation.
The explosion ttbe is made of Pyrex glass, 2 inches I.D. and 48 inches in length. (Corning Class Works offers such glass pipe as a standard item.) The L
L ignition source at the lower end of the pipe is either: 1) an electric are applied across electrodes that are spaced approximately 1/4 iach apart (a standard labor-j story. Tesla coil (MO KV) is suitable); or 2) an electric squib (Atlas Electric Match, Atlas Powder Co.).
y\\
The mixing pump can be any type of centrifugal or positive displacement pump that is capable of circulating gases at the rate of approximately 1 cubic ft/ min or greator. It should be capable of maintaining a vacuum and withetsading
~
several atmospheres of pressure. In order to check the circulation efficiency of
]
the r 'smp, a small amount of smoke can be admitted to the' lower section of the glass tube and the time required for uniform dispersal cf the smoke can be measured.
~
A time period of the order of one minute is satisfactory. D e pump then should be operated for several minutes to insure thorough gas mixing.
{
The large exhaust valve attached to the 1-1/2 inch pipe-cross at the lover end of the glass rube should be an approximately 1-1/4 inch ball valve.
l The valve designated "A" in Figure 3.1, and referrod to as the second 3
valve beyond the desiccant tube, separates the test section from the manifold section. Standard harhare can be used in both sections for valves, tubing, and
,j fittings. EcVever, if the line cor.necting the mixing pump with the explosion tube is 1/2 inch diameter or greater, rapid mixing vill be promoted. It is also recom-l mended that the valves in this line be 1/2 inch ball type. All other lines can i
be 1/4-inch, except in the H O saturator section (see below).
2 The vacuum pump should be capable of producing a vacuum of u 1 mm Hg.
An ordinary U-tube manometer with a scale that reads 0 to 760 mm Hg pressure.
j graduated in 1 na units, is adequate. The desiccant tube is a 2-foot length of j
3/4 inch pipe, filled with a drying agent such as CaSO. The compressed test
{
4 t
Coward, H.F. and Jones, C.W., " Limits of Flammability of Cases and Vspors,"
Bulletin 503, U.S. Bureau of Mines (1952) i 3-8 20<
I' L'
6
(*
gases, which should be of high purity, are obtained from a coursercial gas supplier.
l and their co*cposition should be certified.
i For testing at elevated pressure the glass tube is replaced with a steel tube of the same dimensions (1/4 inch thick tube wall is satisfactory). The pipe-
' ~ -
cross at the lower end of the tube is replaced with a high-pressure steel tee-fitting. No 1-1/4 inch bal14 valve is required since the high-pressure testing is g
conducted at constant vo'lume rather than constant pressure. A 0-50 psia Heise g.
Sage, or equivalent, is used to measure pressures. Thermocouples (ChromeldAlumel) b are inserted into the steel tube at the top and mid-paint.
hese serve to detect flame propagation as well as to monitor temperature for elevated-temperature tests.
b The steel as well as the glass tube, and also the connecting tubing, are wrapped with heating tape to allow elevated temperature testing.
A H O saturator can be added to the system if it is desired to conduct 2
tests with the gases saturated with H O vapor. This saturator consists simply
.g 2
b.
of a bubbler tube, as shown in Figure 3.1.
Calibrations traceable to NES are performed on all measuring devices such as gages and thermocouples. Certifications are obtained for all test gases.
3.4 PROCEDURE FOR CONDUCTING EXPLOSION-LIMIT HEASUREMENTS The attached Figure 3.1 represents a diagram of the apparatus that is
(,
employed for the measurement of explosion limits of H -N /0 -Hal n 1301 nixtures.
2 2 2 The following procedure is to -be followed when such measurements are performed.
1.
The entire apparatus is evacuated, using a vacuum pu=p, to 'zero pressure reading on the U-tube mercury msnometer. The Heise gage. H O saturator, 2
thermocouples, and heeting tape are not utilized when conducting tests at ambient
~
temperature and pressure. The use of these special items is described separately below. If ignition is to be by electric squib, a fresh squib must be attached to the igniter leads prior to pump down.
2.
With the vacuum pump cut out of the system, the test gases are J.
admitted one at a time to the test section of the systen, beyond the second valve following the desiccant tube. Standard technique for preparing gas eiztures by the partial pressure method is followed:
L p
3-9 M<
l j
f
a o
l J
i I
- i l L.
t a.
It e desired volumetric composition of the mixture is pre-determined, and the cor' esponding partial pressures of each r
component of this mixture are computed.
b.
ne first test gas is admitted to the system to the re-I qt.ited partial pressure of that congonent. Ths manifold sec-g
' li tion of the system is then reevacuated (up to the second valve beyond the desiccant tube). The second component gas is then 7.,
{
admitted to its required partial pressure. The evacuation is repeated and the third component is admitted in the same manner as the first and second. Following admission of the 6
third gas the total pressure in the test section of the system should equal or slightly exceed ambient c.
The order in which the gases are admitted is not critical.
}
although, since there is a wide dif forence in their densities I
and since they are introduced at the lower end of the system.
~~
I the order 1201-ll 'N /02 "i11 P' * **1I'"i*i"8*
t-2 2 L.
3.
With the accond valve following the desiccant tube closed, the gases are mixed fer at least three minutes with the mixing pump.
4.
ne nizing pu=p is turned of f and the large exhaust valve at the lower end of the glass explosion tube is opened.
5.
Ignition is sceomplished iranadiately after opening the exhaust L
valve by either firing an electric squib, or by producing an electric discharge across approximately a 1/4 inch gap between ciectrodes using a high voltage
[
transformer (a laboratory Tesla coil is suitabic).
6.
If a flane propagates at least half way to the top of the glass tube the mixture is considered explosive. When a flarw self-extinguishes before k
reaching the mid-point, or does not propagate at all, the mixture is considered non-erplosivo. Ibubtful cases will be considered explosive.
7.
All data and results are entered in a standard laboratory note-book, which is than signed and witnessed.
3-10
O J
(.
8.
An explosion-limit diagram vill be mapped for each set of test gases and conditions. A sample diagram for the H -air-1301 system is shwn 2
in Figure 3.2.
Since the " peak percentage" is the most important feature of the diagram, care vill be taken to insure that an adequi te number or' explosive and nou-explosive mixtures are tested in this region in order to define the peak point.
E' 9.
Electric power is applied to the henting tape to conduct tests at elevated temperature. The apparatus is heated an'd stabilized prior to testing.
~
Temperature is monitored by means of the thermocouples.
10.
In order to perform tests with gases saturated with H 0 vapor.
2 the H O saturator by-pass is opened and a prepared gas mixture is circulated 2
thrcugh it by means of the mixing puur. Cases are judged to be saturated when condensation readily occurs on the walls of the explosion tube.
11.
The same general procedure is used for tests at elevated pressures as at ambient, except that gas pressures are monitored with the Heise gage instead u
of the manometer, and no valve is opened prior to ignition (the tests are per-formed at constant volume rather than constant pressura). Response of the thermo-couples are used to judge flame propagation in the steel tube. Prior to ignition the valves at the upper and lower ends of the explosion tube are closed, thus isolating the explosion from the manifold, measuring and mixing section of the apparatus.
L g
I, e
l 9
i, l
f e
a
v l
r--
.~
0-50 psis Heise Gauge
_L
\\
Thermocouple.E(
I b
O 2 in. ID X 48 In.
Glass Emplosion*
h HO e-2 Tube Saturator
~
Thermocouple %
Mining Pun.p v
y Mercury C
,))
[ Menometer N
g N*
)
c
^
cy
)
ticatingTape 4
(
j Exhaust
+
(
O -N 1301 H2 Air 2 2 y p rp Exhaust e
mm
(-
A Desiccant b Igniter Figure 3.1. Schematic of Apparatus for H2 Explosion Limit Measurements.
I i
1 I
e
q s
1 I
l b-1301 u
A
/xs QNo Explosion 80 70*F g
GExplosion Propagated
\\
l atm t.,
8 Partial Propagation
/
Dry Gases 5
Squib ignition
/'# : '60-NNhp o
t.
~ /\\/WWV\\
o 1
/\\/\\/VV
/\\/\\/\\/V
/\\
/\\/V N/\\/N;K N h/\\
i
/NA/MK/I! VNVNA iHZ
,'t Figure 3.2. Explosion Umits. H. Atr.1301.
2 3-13 I
\\
\\}
25<
_~
O-
\\,,,
Ii
(,;
SIM1ARY OF TEST RESULTS EXPIDSION - 1.IMIT HEASUREMENTS
{
(To accompany Figure 3.2) t (70*P,1 atm, dry gases, squib ignition) f l
yolumetrie 7.
i i I
("
H2 A,17, Result Test No, g
4-1 29.3 12.3 58.4 No explosion l
4-2 "
19.7 14.0 66.3 Explosion propagated
)
5-1 21.7 13.8 64.5 Explosion propagated 5-2 23.7 13.0 63.3 Partial propagation 6-3 24.5 12.4 63.1 No explosion I.
7-1 0
3.6 96.4 No explosion
+
7-2 0
5.9 94.1 Explosion propagated 8-4 0
71.2 28.8 No explosion 8-5 0
70.7 29.3 No explosion 8-6 19.6 22.5 58.0 Explosion propagated I
9-7 23.0 21.6 55.4 No expleaton 9-8 20.9 22.0 57.0 Explosion propagated j
99 22.4 21.3 56.3 No explosion 10-10 17.7 34.6 47.7 No explosion 10-11 12.9 37.1 50.0 No explosion 11-1 8.8 36.3 55.0 Explosion propagated 11 2 12.3 34.1 53.6 Explosion propagated
~
11-3' 15.0 34.2 50.8 No explosion 11-4 0
63.7 36.3 No explosion i"
12-5 14.6 8.3 77.1 No explosion 12-6 9.7 10.1 80.2 Explosion propagated l
($
12-7 24.2 16.5 59.3 No explosion 9'-7 15.0 9.0 76.0 No explosion f,
10'-1 14.9 10.1 75.0 Explosion propagated 10'-2 10.0 9.1 80.9 No explosion
[
28<
3-14
_~
i
-l
~
._t
[.
i 1
1301
'f g
80 G No Explosion
/\\
7L* F
]
G Explosion Propagated
/
\\ \\
1 atm
\\ [A Ory Gases gg
\\
9, Spark ignition s
/N;e:'%g ?
- l 1
~/\\/\\/\\/\\/\\/\\
AA/VVW\\
l A/\\/\\/VN/MNA
/\\/\\/\\/\\/\\ W N N \\
AI' H2
- l
~
n
~
I Figure 3.3. Explosion Limits, H. Air 1301.
2 3-15
\\L z?:
)
SUMMARf 0F IIST RESULTS i
EKFLOSION - 1.IMIT 2EASUREMEWS l g-(To accompany Figure 3 3) 1 L i
(70*F, 1 atm, dry gases, spark ignition)
Volumetric 7.
Test No.
130i 3
Air Result 13-1 0
73.2 26.8 Explosion propagated
}
L3-2 0
75.5 2's.5 No explosion 13 -3 24.4 12.6 62.9 No explosion
~
13-4 21.6 13.6 64.8 No explosion 1
14-5 19.7 14.0 66.3 Explosion propagated 3
14-6 21.7 17.1 61.2 Explosion propagated
,i 14 -7 23.2 17.1 59.8 No explosion 14-8 21.9 14.3 63.9 No explosion 14-9 l
9.9 10.7 79.4 Explosion propagated 14-10,
13.6 9.6 76.8 Explosion propagated 15-11 15.2 9.1 75.7 No explosion 16-1 22.0 22.3 55.8 No explosion
]
16-2 20.0 22.1 57.9 Explosion propagated U
16-3 23.1 22.4 54.5 No explosion 16-4 18.0 34.8 47.3 No explosion d
16-5 14.1 35.0 50.9 No explosion 17-6 12.2 35.3 52.6 No explosion
]
17-7 10.6 35.0 54.4 Explosion propagated 17-8 7.0 49.7 43.3 No explosion 17-9 5.0 50.7 44.4 No explosion 17-10 3.1 51.0 45.8 Explosion propagated l
28<
g.
3-16 "d
r-,,
r-
w u
I i
1301
~
t:
/\\KN O No Explosion 80 120*F GExplosion Propagated I atm 6 Partial Propagation
\\
Dry Gases l$
Squib Ignition
/'W%/\\
/\\/\\/\\/\\ V \\
~
/\\/\\/V
, N\\
/\\/\\/\\/1'k/V4/\\
/\\/\\/Wetl\\A/%/\\
L
/\\M/\\/\\/V\\XN/\\%
Spark ition j-Figure 3.4. Explosion Limits, H.N /0,2/1 1301.
2 2 2
{
3-17 t
23<
l l
p
iL i
t
SUMMARY
OF TEST RESUI.3 EXPI4SION - I.IMIT MEASURN (To accompany Figure ". 4) i (120*F,1 atm, dry gases, squib ig:ition)
Volumetrie 7 4
j Test No.
ge H2 2:1. 3 42 Result j
2 15'.0 26.7 No explosion 50-1 58.3 50-2 57.2 15.6 27.2 No explosion 51-3 55.1 16.5 28.4 No explosion G
51-4 50.3 19.7 30.0 No explosion r*
52-7 28.7 33.3 38.0 Explosion propagated 52-8 28.2 8.1 63.7 No' explosion 1
55-1 44.3 10.8 44.9 No explo.sion 55-2 42.8 11.6 45.6 Explosion propagated l
q j
56-3 43.3 17.0 39.7 No explosion 56-4 40.1 20.0 39.9 No explosion l
57-5 40.1 22.1 37.8 No explosion i
57-6 40.1 10.0 49.9 No explosion 1,
58.7 40.2 12.0 47.8 No explos' ion 59-1 44.2 18.1 37.8 No explosion 59-2 35.3 24.9 39.8 Explosion propagated t
u 60-3 31.3 26.8 41.9 Explosion propagated 60-4 35.2
- 9. 8 35".0 No explosion b
61-5 28.3 36.8 34.9 No crplosion a
61-6 28.0 11.0 61.0 Explosion propagated 62-5 40.3 14.9 44.8 Explosion propagated
'~
62-6 42.3 13.9 43,3 Explosion propagated' 63-7 45.5 12.3 42.2 No explosion
{
i i
63-8 43.9 14.9 41.2 Explosion (partial) j s
64-9 44.4 12.8 42.8 No explosion 53-9 85.1 14.9 Explosion propagated 53-10 86.8 13.2 Explosion propagated
^
54-11.
89.0 11.0 No explosion J'-1 25.0 7.9 67.1 No explosion
+
1 l
6-l 30<
i I
3-12 5,
.-m--
['
J-i Test fM.
1901 H2 2:1. N2/02 Renuit 2
9'-2 25.0 9.0 66.0 No explosion 9'-3 25.0 10.0 65.0 No explosion 9'-4 25.0 11.1 63.9 Explosion propagated 9'-5 15.0 8.1 76.9 No explosion t.
9'-6 15.0 10.0 75.0 Explosion propagated l
l.
l.-
i e
e I
(.
i'
(.
p.
1.
4 6
I i.
31<
i 3-19 b
_g
l l l
,,1
,4 80 C No Explosion 70* F O Explosion Propagated.
/
i 1 atm
/
par I ition
/'#:%/\\
a A/\\/\\/RN\\
~
a A/\\/VNf%\\A l
A/\\/\\/M/'VW\\
l
/\\/\\/\\AQANA/NA
/\\L\\Af\\/O\\!AM\\/\\X\\
2
-- N /02 2
Figure 3.5. Explosion Limits, H;-N,07 2/1-1301.
3 I
3-20 t-I-
U2<
i i
i I
,J SUHHARY OF TEST RESULTS f,
1 EXPIDSION - LIMIT MEASURDiENTS (To accompany Figure 3 5)
(70*F,1 atm, dry gases, spark ignition)
VOIINETRIC 7 i
0 Test No.
- 1301 H2 2:1. N2/ 2 Result
]
65-1 38.1 15.2 46.7 No explosion 65-2 36.3 15.8 47.9 Explosion propagated j
]
66-3 37.4 13.8 48.8 No explosion 66-4 37.4 17.7 44.9 Explosion propagated 67 5 37.1 20.1 42.8 Explosion propagated 1
]
sJ 67-6 39.3 16.9 43.8 Explosion propagated 68-7 37.1 23.1 39.8 No explosion 68-8 41.0 17.0 42.0 No explosion 88.9 11.1 Explosion propagated 69-9 i
69-10 90.0 10.0 No er.plosion
'l
~.
~
it t
33<
l.
3-21 i
$r equ
=*
- 7---'
f P
- ~
~. _.. -..
h 1301
/\\"
GNo Exploskn 8,0 70*F OExplosion Propagated u\\
l atm O Partial Propagation
,z J Dry Gases jO Squib ignition 3
App &/\\
~
/\\/VW\\'vhA
~
/\\/\\
/WA/\\
i
/\\/\\/' /NMVVN i
/VNAN/,U"&\\/YN i
l
/\\/}WNA/\\/\\A/\\/\\L N,
Figure 3.6. Explosion Limits. H.1:1, N iO
- 1301, 2
2 2 t
i 3-22 l-'
34<
a
. - - ~...
= -
l, en i
SUMMARY
OF TEST RESULTS EXPIDS10N - LIMIT MEASUREMENTS (To accompany Figure 3.6)
(70*F,1 atm, dry gases, squib ignition) 4 Volumetric 7 i,
Test No.
1301 H2 1:1. N2/02 Result I
29-3 54.7 16.1 29.2 Explosion propagated I"
29 4 56.5 15.3 28.2 Explosion propagated 30-5 58.3 14.8 26.9 No explosion d
30-6 59.5 14.2 26.3 No explosion 31-7 60.I 14.1 25.8 No explosion I
.J 31-8 58.5 17.8 23.7 No explosion l_
32-1 58.2 11.8 30.0 No explosion
! ]
32-2 54.2 21.8 24.0 No explosion 33-3 53.9 13.2 32.9 Explosion propagated 33 4 57.3 16.9 25.8 Explosion propagated 34-5 53.0 20.0 27.0 Explosion propagated 1
35-1 53.4 10.9 35.6 No explosion J
4*-1 44.9 10.1 45.0 No explosion 4'-2 44.9 11.1 44.0 Explosion propagated lj 4'-3 35.0 10.1 54.9 No explosion 4' 4 35.0 11.0 54.0 Partial propagation f.
4'-5 35.0 12.0 53.0 Explosion propagated A'-6 25.0 10.0 65.0 Explosion propagated 4'-7 25.0 9.0 66.0 Explosion propagated 5'-3 25.0 8.0 67.0 No explosion 5'-9 15.0 7.0 78.0 No explosion I
5'-10 15.0 8.0 77.0 Explosion propagated 5'-11 4.0 96.0 No explosion 5'-12 5.0 95.0 Explosion propagated 5'-13 6.0 94.0 Explosion propagated j
5'-14 7.0 93.0 Explosion propagated 1
35<
I; 3-23 f
L.
Test No.
g 3
til. N 0
2/ 2 Resu13 t.
5'-15 45.0 29.9 25.1 Partial propa8ation j
6'-16 45.0 29.0 26.0 Partial propagation 6'-1 45.0 28.0 27.0 Explosion propagated 6'-2 45.0 31.1 23.9 No explosion 6'-3 30.0 48.0 22.0 No explosion j
6'-4 30.0
- 46.0 24.0 No explosion 7'-5 30.0 45.0 25.0 No explosion j
~
7'-6 30.0 44.0 26.0 Explosion propagated f
7'-7 15.0 60.0 25.0 Explosion propagated
["
7'-8 15.0 62.0 23.0 Explosion propagated 7'-9 15.0 64.0 21.0 Explosion propagated I
J 8'-1 15.0 80.0 5.0 No explosion 8'-2 15.0 75.0 10.0 No explosion
,j,
(
8'-3 13.0 70.0 15.0 No explosion 8'-4 15.0
- 67.0 18.0 No explosion 8'-5 15.0 65.5 19.5 No explosion L
8'-6 90.0 10.0 No explosion e
8'-7 88.0 12.0 No explosion l
' l 8'-8 86.0 14.0 Explosion propagated U
=
li i..e i4 3
9 1
0 1 24
=..-.
4 y
-g pr,-e-um-y ye -,.-
e.
4P
Q
'!}
w I
C n
1301
~
c y\\
l G No Explosion to
\\ (!
p O Explosion Propagated et b
OPartla! Propagation J
/\\/ifi/\\00;"-
/\\/\\/ho/\\
o
/\\/\\AFM/\\
l
/\\/\\/X/\\/ix/k'\\
/\\/NA/\\/\\V\\/\\/\\
I; l!,,/@VN/\\/\\/'\\/\\/\\YX/\\/\\/\\/\\ d\\/\\
/\\/
1 Figure 3.7. L$csion UrJts, H *I:I N '02 1301.
2 2
3-3
{'
37<
l 4
m SUMrfARY OF ':ZST RESULT 3 EKPIDSION - LIMIT MEASURDOWIS (to accompany Pigure 3.7)
(70*F,1 atm, dry gases, spark ignition)
!.t IU
- Volt:secrie %
1 e
=
fb Test No.
1301
,1 1:1. N2/ 2 Result 0
19-1 47.7 16.7 35.7 Explosion propagated 19-2 31.3 18.7 29.5 Partial propagation 19-3 52.7 16.7 30.6 Explosion propagated ij{
19-4 53.7 15.4 30.9 No explosion d
20-1 54.5 17.6 27.8 No explosion 20-2 50.8 24.8 23.7 No explosion 21-3 48.3 24.9 26.8 No explosion 21-4 50.0 13.7 36.3 No explosion g
,j 22-5 49.4 24.6 26.0 No explosion j
23-1 45.0 29.9 25.1 No explosion
{
23-2 44.1 27.8 28.1 Explosion propagated 24-3 '
48.7 22.3 29.0 Explosion propagated
! C 24-4 30.4 9.8 L 9. 8 No explosion
,[
l i
25-5 30.4 13.2 56.4 Explosion propaga ed
?
j 25-6 24.0 58.1 17.9 No explosion
{.
26-7 23.9 55.4 20.7 No explosion l_
26-8 23.7 51.5 24.7 Explosien propagated j
27-9 0
86.5 13.,5 Explosion propagated 27-10 0
86 ;
11.3 Explosion propagated 1
j
'l 28-1 0
93.8 6.2 No explosion 28-2 46.2 31.8 22.0 No explosion 1
.J e
3-26 m
O*
IgI i
(
I l
1301
~
id
/\\!/\\
C No Explosion 80 120*F di g I atm
~ O Explosion Propagated i ;
lL A Partlal Propagation
.~n= I
\\
Dry Gases
/\\/c 8
1,0 Squib ignition Near Peak s!
\\
spara ignition to-or pow:::
j j
- u
/\\/\\ W iu
/\\/\\/8/\\ \\A
~
/ # :/ X / M N / h / \\
A/\\#NVNYN/\\A
!i
/\\/\\A/\\AinM/XA
/\\pv\\/\\/N!/\\AN/\\/Nh lit ii'
/N 'Ill*I30I*
Figure 3.8. Explose Umits Hd2 2
3-27
!]
39<
y-
4 I
~
SUMMARY
OF TEST P2SULT3 j
EXPLOSION - LIMIT MEASURDfE.'fTS J
(To accotopany figura 3.8)
(
(120*F,1 atu, dry gases, squib ignition)
V,,9}y.ge t rie %
N
,,,3, lit. N2/02 Result Test No.
12p,,1, 36-1 58.3 15.0 26.7 No explosion j
36-2..
57.0 15,3 27.7 Explosion propagated i
37-3 57.4 12.9 29.7 No explosion
.}
37-4 57.0 17.1 25.9 Partial propagation
't 38-5 55.5 19.7 24.8 No explosion 39-1 57.6 18.4 24.0 No explosion u;
39-2 55.4 11.8 32.8 No explosion 40 3 55.4 13.7 30.9 Explosion propagated 40-4 55.4 21.8 22.8 No explosion 41-5 50.2 24.1 25.7 Explosion propagated 47-8 50.1 13.0 36.9 Explosian propagated 48-9 50.1 10.9 39.0 No explosion j
48-10 50.1 26.1 23.8 No explosion 49-11 49.0 29.0 22.0 No explosion Spark Trnittnn 44-1 99.1 9.9 tio explonion 44-2 88.0 12.0 Explosion props. gated 45-3 25.4 54.7 19.9 No explosion 45 4 26.0 49.2 24.8
!!o explosion 46-5 25.2 44.9 29.9 Explosion propagated 46-6 30.4 10.0 59.6 No explosion 47 7 30.1 13.4 56.5 Explostoi propsgated 49-12 25.2 48.1 26.7 Explosion propagated I.
l 40<
3-28
-m-
.w-m.
--.w.
m,,
n-wn--
-m,
..n.-,.
m-
,,n~-,-
p B
i a
4 L
g 1301 L
A
/\\#\\
l u
- ==.... A/VN i
c N\\ M/\\Cr
[";f'fxM L
N \\/\\W R /\\
~
t N \\/\\M AN N \\
L N\\/\\iUNy\\NN c
I A/\\/\\A/\\N\\/Ny\\
N /02 GI 2
\\
j H2 Hp
~
Figure 3.9. Explosion Umits H.0 /N + I/I
- I301-2 2 2 3-29 41<
l'
i l
i o'
~
f ld SUMARY OF ' TEST RESUL'iS EK21DS10N - LIMIT FJ%St,7EMENTS (to accompany Figure 3.9)
(120'F,1 atm, H O maturation, squib ignition) 2 Volumetrie 7.
a
,.I V
Test No.
1301 H2 1:1. N2/02 Result j
70-1 47.3 13.8 27.4 Explosion propagated
]
70-2 48.9 13.6 26.3 Explosion propagated 70-3 51.8 13.5 24.4 No explosion J
71-4 51.2 14.8 22.7 No explosion 71-5 45.5 15.9 27.2 Explosion propagated 71-6 48.3 17.7 22.7 No explosion 72-7 45.9 10.7 32.1 No explosior.
72-8 43.8 13.3 31.5 Explosion propagated 72-9 50.8 11.9 25.8 No explosion 72-10 46.6 17.1 24.8 Explosion propagated 72-11' 26.6 12.8 49.1 Explosion propagated i-73-12 26.6 9.6 52.3 Explosion propagated l ll 73-13 27.4 8.6 52.5 No explosion 73-14 22.9 42.3 23.2 Explosion propagated j
73-15 23.1 44.7 20.6 No explosion 76.2 12.2 No explosion-73-16 74.6 13.7 Explosion propagated 73-17 I
I lt
!I t
I l
$ s l~'
42<
4 3-30
-l lI i.-
2 f
1301 O
k u
fy, l
120* F C No Explosion g +.
'.}.
O Explosion Propagated
,01 18 3atm I
Dry Gases I,
Squib ignition i
Nonexplosive R
\\egion, xW/\\
/
!i;
/\\/
N/\\/
i;
/\\/ XIA/X l;
/VVN/%f\\/\\
/\\/X/N/\\AX/N/\\
/\\j>(/\\/VVNAN/\\/\\b N /Og 2
2 f
Figure 3.10. Explosion Limits H -0 /N 'I/I
- I30l*
2 2 2 3-31 4a<
n.
- il J
'l 1
i i}
SLHMARY OF TEST RESULIS
(
II7IDSION - LIMIT MEASURDfENTS
{.
j (To accompany Tigare 3.10)
~
(120*7, 3 at:n, dry gases, squib ignition)
Volumetric *.
g Ls: No.
1* 5 R2 If1. N 0
2/ 2 P.nsult 74-1 56.6 -
15.4 28.1 Explosion propagated u
. 74 55.7-4LS 30.2 Explosion propagated 74-3 58.G 19.8 24.3 Explosion propagated
~
74 -4 58.3 18.0 24.0 Explosion propae,ated 75-5 55.3 25.0 20.0 No explosion 75-(
60.3 15.7 24.3 Explosion propagated 75-7 63.0 12.3 24.7 No explosion l'-1 60.9 10.0 28.9 No explosion 1'-2 62.3 17.5 20.5 No explosion l'-3 55.0 10.0 35.0 No explosion l'-4 90.9 11.8 27.5 Explosion propagated l'-5 60.0 17.7 22.3 Explosion propagated l'-6 30,2 10.0 61.0 Explosion propagated 2'-7 30.3 7.0 63.0 No explosion 2'-8 40.0 40.0 20.0 Explosion propagated 2'-9 40.7 43.0 17.0 No explosien
,h 2'-10 90.0 10.0 No explosion 2'-11 87.7 12.3 Explosion propagated
{
2'-12
..) 3 58.0 17.0 No explosion 3'-1 25.0 55.0 20.0 Explosion propagated
.1'-2 5.0 95.0 Explosivn propagated l'
q?
4.0 9'.0 No explosion 3'-4 7.0 93.0 Explosion propagated 10*..
/ '
..'s,
10.0 50.0 Explosion propaasted 10'-2 40.7 9.0 41.2 1.xplosion propagated 10'-3 40.C 8.0 52.2 Explosion propagated l
11'-1 40.0 7.2 52.8 Partial propagation I
11'-2 40.0 6.0 54.0 No explosion 11'-J 20.0 5.9 74.1 No explosion 11'-4
- 20. *,
7.0 73.0 No explosion 11'-5 20.0 8.0 72.0 Explosion per.pagated bb 3 32 y-e--m e
v v
1 J
l' 6
1:
I i
\\
11 i
n.
5 i
t 4
L.
l g
t
~
4
--> Air u.
Squib 4
Ignition
)
,E3
~
z
\\
Spark ignition ->
\\
\\
,1 s
\\
\\
\\.
l
\\
\\ \\.,{ 3 atm j
H O Vapor 8% C 2
\\
Saturated
\\
Ii l
0 0
20 40 60 80 HALON 1301 PEAK PERCENTAGE i
I Figure 3.11. Halon 1301 Sek Percentate for Supproulon of Explosions of 2
Mistures (70*F: 1 atm).
Hydroger, sn'. 0 /N2 3-33 45<
.~.-___..
_U
~"
........__.L
p.
4- - as.
- r--
1 i
i
'w I'
4 u
l 3.0 4
,5 9
i I
a l
P' 2.0 a
n.
S I
d u
y w 1.5 g
]l.;
R e
ae
,l 1.0 i
e
'n g
j 0.5
,.)
J
]
O u
54 55 56 57 58 59 60 PERCENTAGE H ALori 1301 1
'.m Figure 3.12. Plot Shewing !Asasured Pressure increase When Mixtures of H2 and r,
1:l/Ng:0, Containing Halon 1301 were ignited in a 5.6 cu. ft i
Sphencti Veuel.
I 3-34 I
46<
k.
P 4
~
\\
O i
l I
l i
4.0 LARCE-SCALE TESTING _
u 4.1 ENV1SIONED PROBLEMS I
This task of the progran involved the fabrication of a system that t: L essentially simulated on a sub-scale a containment vessel containing a Halon 1301 inerting system. The objective of the task was to inject Halon 1301 into a large test chamber under various conditiens similar to those that might be present in a containment. Any problems that might develop could be identified and corrected. To illustrate that problems were conceivable the fellowing ex-
~
(l amples are presented.
L A containment vessel of 34,000 cubic foot ullage might require of the order of 20,000 pounds of Halon 1301 for inertirig. This Halon would be in-j jected as a liquified gas which would have to vaporize in the containment vessel. The heat of vaporization of 1301 at 75'F is approximately 35 EIU/lb.
Hence, 7 x 10 BTU of heat would have to be supplied from some source to vaporize the Halon. If the containment vessel contains 7 psia of air at 120*F y
at the time of a LOCA (taking no credit for steam), the total weight of air present is 1100 pounds. Using a specific heat of 0.24 DTU/lb'F a value of 264 Bid /*T is calculated for the air. This means that since the extraction r
I t L.
of 264 BTU drops the air temperature one degree, it is impossible for the con-5
,f{l tainment air alone to supply the necessary 7 x 10 BTU needed to vaporize the
~l u 1301. There is ample heat in the containment vessel hardware, so the caly l {;
question is whether any heat transfer problems might exist. This was one of L
the questions to be addressed in the large-scale testing.
There are other conceivable problems connected with 1301 injection.
The Halon would be discharged through nozzles, which together with the pressure i
determine the discharge rate. The pressure at the nozzle n.ast exceed the vapor pressure of the Halon to insure that the agent is injected as a liquid.
r
)
If the nozzle pressure is below the vapor pressure (229 psia at 75'F) then This f
(theoretically) the agent can vaporize prior to issuing from the nozzle.
premature vaporization could cool the piping and possibly cause freezing of i
the Hr.lon 1301 and atoppage of the flow. Ordinarily this would not occur with a properly designed aystem. Ecwever, if a nozzle ruptured, the flow rate might increase enough to cause the discharge pressure to drop below the vapor i
pressure value. Thir could possibly lead to the above-described probles.
4-1 4'/<
)
I
~
b i
t, L
A similar problem might occut if the Halon piping became over-
{
heated. Prsaature vaporization in the pipeline at the ho zone could lead to freezing and stoppage of the flow. In one of the tests to be described below j
it was found that under certain conditions the flow could be stopped by heating L-the line carrying Halon.
In addition to exploring possible problem areas, the large-scale testing was intended to provide information for design of an actual system.
Also, large-scale ignition tests were performed tc verify the smaller-scale explosion-limit results. In these latter e v'.e sixturas of 1301, H, 0 and 2
2 H were prepared in the 1200 cubic foot test usel an.1 igniters fired in the 2
gases. A description of the ignition tests was given in Section 3.2.
4.2 DESCRIPTION
OF TEST FACILITY The large-scale testing was conducted in a 1200 cubic foot steel vessel approximately 9 feet in diameter by 18 feet in length. The vessel had an access door, several viewing ports, and provisions for piping and electrical l
1eads. A schematic drawing of the test facility is shown in Figure 4.1.
We drawing is fairly self-explanatory but some specific itesa can be described.
The Halen was purchased from DuPont in 2000 pound lots. Ihe Halon storage tanks were pressurized with N2 gas when required, and the 1301 was always drawn as liquid from the bottom portion of the tank. Usually a delivery pressure of 360 psia was used. The liquified Halon flowed through a high-i.
pressure flowester where delivery rates were measured (Wallace and Tierman Model 5220M42008 " rotameter-type" flowmeter, custom designed for the project).
This allowed flows up to approximately 33 lbs/ min of Raion 1301 to be measured.
It was found that upon opening the valves to initiate Halen flow, the system functioned properly with no problems. Under one set of conditions (no N 2
pressurization, and heated lines) some difficulty occurred which is described in the following Results section.
5 Pressure measurements in the tank vere made with a precision Reise gage. Thermometers and thermocouples were used as required for temperature meas urements. All instrumentation was calibrated per Quality Assurance re-quirements, traceable to NBS. One-half inch diameter stainless steel lines l
were used for piping the Halon and other gases. Hydrogen and oxygen were i
4-2 i
48<
l 1
-... ~.. -. -.
-...w l
.~7
.7 T
~~
~
IC C'
i--~
(C I' ~ '
C LJ VACUUM PUMP PRECISION 0-50 psla A
PRESSURE GAGE W
THERMOCOUPLE ItYDROGEN AND SPRAY OXYGEN SUPPLY f NOZZLES CAMERA T
g e
A f
1200 f t3 f
[l TEST CHAMBER a
1 r
H O
1000cfm PERSONNEL y
2 c.
O CIRCULATING FAN ACCESS DOOR N OW n
LIQUID flALON VAPOR FLOWMETER b
2000 lbs WElGHT g
HALON 1301 SUPPLY VENT LIQUID i
U2
~( N11 ROGEN PRESSURIZATION TANK b
Figure 4.1. Schematic Diagram of Test Facility Used to Conduct Large-Scale j
Halon 1301 Vaporization and Ignition Tests.
i I
u
. ~.. -
O
l l
i i
<}
J, admitted into the tank as ganas as required for the ignition tests. Gas composi-tions were datermined by partial pressures.
he spray nozzles empicyed in the tests were coemarcially availab'le I
types (purchased from Sprayins Systems Co.).
The nozzles were chosea to pro-j vide two different spray patterns - full cone and flat spray. It was thought j
initially that it vould be important to have the liquified Halon discharge in such a way that it would make good contact with the vessel walls in order to promote efficient vaporizatics. It also seemed that the spray pattern of the d
noaala might have a significa=t effect on this. It was found, however, that a
either apcay pattern produced enormous turbulence which scattered the fog-like l
liquified Halon continuously thrceghout the test chamber. As will be seen in a
the naze section the rate of vaporization followed closely the injection rate.
El The rate of injectice of Ratan is determined by the nozzle orifice siza and by the discharge pres. aura (which was essentially the storage tank pres-
'i j
sure of 360 psia since practically no pressure drop occurred thvough the piping).
The foll.owing tabulation gives this injection rates at 360 psia storage pressure for the nozzles employed.
Number Halon 1301 Spray Systems Spray Equivalent Nozzles Injection Co. N> del No.
Pat te rn Orifice Diameter Used Rate 11C04 Flat 0.052" 2
20 lbs/ min 3
ij 1/4 TC-2 Full Cone 0.047"
~2 13 lbs/ min TU-SS-5 Full Ccne 0.0 82" 2
33 lbs/ min
. ~i
.i The ayaten design was based large.ly on data provided in the National Fire Protec -
1 tico Association, Standard 12.1, *Ialogenated Extinguist Ing Agents Systens, Halen 1301."
l
]-
The question of hcw to scale the large-scale test results should be l
'J ccmaide red. This question =1-ty relates to whether Halen vaporization vill
!g occur readily in a cont =4rw-- vessel as it did in the ARC test chanber. Con-
['
sideration was given to the value of the surf ace area-to-volume ratio of the test chamber ecspared to that of a enntaiwnt vessel. The SA/V ratio of the AIC test chamber was approx 1:.ately 0.53 f t The SA/V ratio of the CNSG con-tri--ent la not so easy to estimate. For dinensions of a cylindrical contain-I~
nent 38 feet in dieciater by 64 feet in height, with 34,000 cubic feet of ullage.
L t
t-4-4
\\-
50<
a I'
J a value for SA/V of 0.29 f t is computed. Since the SA vill be greatly larger
~
then estimated due to the presence of Icactor components, the SA/V ratio of the containment may be very close to that of the A1C chamber. However, it appears
}
upon further analysis that SA/V ratio is not 'the important factor to consider
- ?
in this case as it is in most problems of this type. In answering the question of how should the ARC large-scale data be scaled to a containment application.
the following considerations are presented.
,In the ARC chsaber, Halon was discharged at a rate of about 20 lbs/ min F
in typical tests. As will be seen later, this is roughly the rate required for containment protection. The surface area of the ARC chamber was approximately 2
J 650 ft and total injection time was about 20 minutes. The total surf. ace area in a containment might be of the order of 20,000 ft. The quantity of Halon
]
required vill be in the neighborhood of 20,000 pounds, requiring sn injection time of 1000 minutes. These values are stsanarized below:
r f
N ARC Test CNSC Chamber Containment 1
Injection Rate
. 20 lbs/ min 20 lbs/ min 2
2 J
Surface Area 650 ft 20,000 fe j
I Quantity Halon 400 lbs 20,000 lbs Injection Time 20 min 1000 min There are several cays to consider these data. They are all independant of the number of nozzles empsyed. The quantity of Halon required in the contain-ment is 30 times greater than that in the ARC tacts. Howaver, there is 50 times more time available to discharge, and there is 30 times more surf ace area avail-2 able. Alternatively, the containment providas i f t of surf ace area per pound of Halon 1301; the ARC chamber provided 1.6 f t /lb. As can be seen, the results l
can be scaled very well. The requirement is sonetthat more severe in the contain-
~
!~
ment but is still of the same order of magnitude. The closeness of the require-ments in the ARC tests and in the containment, and also tha fact that no hint of any vaporization difficulty was encounterad in the ARC tests, provides full assurance that vaporization of Halon in a containment will be no problem.
4.3 RESULTS Approximately 12 large-senle tests were performed, and of these six wara l
regarded as preliminary for f amiliarization purposes and for checking components 4-5
{
51<
\\
\\.s (ib of the system. The coupleta iacord of data was taken on da other six tacts. Ma l
raw data are not presented'is this report, however, records are on file and arail-ll able from Todd or Atlantic lesaarch Corporation.
1 l
Test 1.
Liquified Ealon 1301 was injected into the tank 'at a rata of 20 lb/ min using a naasle that produced a flat sprsy. A circulating fan was oper-ating during the test. The results are plotted in Figure 4.2 in terms of t:me 4
pressure rise in the tank. The solid line represents the rate of pressure rise to be expected if the Halon vaporized instantly as it was injected. The pc.ints d
represent measured values of the pressure throughout the test.
It is seen that the ate of vaporization fel. lows the theoretical (i.e., the injection rate) wery well. The total temperature drop in the chamber was 8'c during injection. The slight drop in the flowuetar reading shoku on the Data Record sheet was eW by a drop in the 1301 storaan pressure from 360 to 340 psia.
Test 2., This was essentially a duplication of test 1 but with the circulating f an turned of f.
Een the data are graphed, a plot identical to Figure 4.2 is obtained. It is apparent by visual observation during the casts that the nozzle discharge creates so much turbulence that fan dispersion of the fog-like mist is not necessary. It is noted that the marimum tes:perature de-crease in the tank of 15'c 4.:ceeded that of test 1.
i!
Test 3.
This test was performed at a reduced flow rate of 13 lb/r'_n, l
l using a nozzle that produce.f a full cone spray pattern. The results are e:ac fd in Figure 4.3.
The theoretical and actual vaporization rates deviate from each i
other by about the same small increment observed at 20 lb/ min injection rs:.em.
Test A.
The highast flow rate measurable with the flowmeter,131h/rin, was used in this test. This rate exceeds the required injection rate for the
.,d i
CNSC containment vessel. The vaporization rates are plotted in Figure 4.4 for i
this test. It was intended to discharge 1301 for a 12-minute period, herewr, i.,
j the storage supply of agent became depleted af ter eight minutes of injectica.
i llowever, it is apparent that the rate of vaporization is very near the thacreti-f cal throughout the test.
i i
Test 5.
In this t.est the 1/2-inch stainless steel piping betwee:n the t
i llalon storage tank and the test chamber was preheated to 200'F prior to f ro
'm ing the 1301 flow. Heat was pr7rided by heating tape and was n.aintsiced thr:r:gte:t the test. The set of nozz*es that provide 13 lb/ min flow rate at 360 psia pressure 4-6 4
52<
~
^
{'
6 3
t 3
hc O
i h
w 2
3 n
~
i t
~
seT s
ro f
k 8
a L
2 a
T e
g mu L
loV 4
e 2
2 gra.
[
Lni nm m
i
/
ebml S
i0 E
0 T 0 i
1 2
U l
)
f 2
I n o L
D im no f
A NWE 3
V O
o PL
(
e O
7 8
D L
E A O J
E i t P
t L
ca ERR F
M nR H
T u
UU S
IT Fa t
6 l
S S
1 E
S ad AE s e at ER c
1 ee 7P sj O
a n ei s r
c a n
T
/
w 2
i 1
e1 r0 us3 s1
~
e W
n r
P o L
gl a n
E iH LS 8
wd
']
o e AIR hi C
Si f
E u
I R U i
m t
T R lPL oq E
i S
OS EE
]
2 HR 4
L 4
/
TP e
r u
6g F
0 0
0 0
0 0
0 0
0 0
0 0
0 0
6 5
4 3
2 1
{Ou uEa.E g 2 5 u e'"
~
gC J
0
~
~.
I
.C
~~
- _.J L.
L _.
~ - '
N E~
L.
l.._
. ' ~ ~ ~ ^ C.
' - ~
'~
~
800 M 'I g
700 y
5 THEORETICAL PRESSURE RISE i
600 e
h MEASURED PRESSURE VALVES f
@ 500 El 3m d
G 1
/
E 400 i
a 2
u A
303 200 f
.)
l su 0
4 8
12 16 20 24 28 32 36 40 TIME (min)
Figure 4.3. Large-Scale Test 3. Plot Showing Pressure Increase es a 17 unction of Tiene in Large Volume Chamber for Test in j
whkh Lt.aulff.J tielun 1301 was injected at Hele of I3 lb/ min.
i
~~'
"*7~
~
- T*
x k
a 8
I I
i f
g I
g I
500 pg }-6f y y w
~
0 CO
?oO O
400 w
c:
3 THEORETICAL PRESSURE RISE E
/*
.-,j g 300
.4 2
1 u
MEASURED PRESSURE VALVES 200
!. $q 1
T
?
100" i
0 4
8 12 16 20 24 TIME (min) 1 i
~
Figure 4.4. Large Scale Test 4. Plot Showing Pressure increase as a f ]l Function of Time in Large Volume Chamber for Test in whkh Liquified Halon 1301 was injected at Rate of 33 lb/ min.
~
i
~
l 4-9 p
55<
1
,3 lI w
.I n
y 1
were used, however, the 1301 storage tank was purposely not pressurized with N2 at tha start cf this tect. The storage tank pressure was equal only to the vapor pressure of 1301 at the ambient temperature, 215 psia at 70*F.
When 1301 flow na initiated, only gaseous agent issued from the nozzles.
I'
)
The gas flow stopped entirely for short periods and then started again. Liquified agent never issued into the test
- chamber. No significani pressure increase was observed in the chamber. This test was repeated three different times with the same resulta each time. Each test issted for several minutes. The difficulty was attributed to evaporative cooling of 1301 in the heated piping which led to freezing and stoppage of the flow.
Following the above attempts to flow 1301 through a heated line, the Balon storage tank was pressurized with N2 gas to a pressure of 360 psia. The test was then repeated with the 200'F line. With the higher pressurization the flow was satisfactory. However, during the first minute the flow usa somewhat trratic, increasing and decreasing in intensity. Liquified 1301 was observed to w
be issuing from the nozzles during the first minute, Af ter one minute the flow was completely smoothed and normal for the remainder of the test. The pressure data of the accon:panying Data Record sheet refer to pressure rise at the. storage 7
pressure of 360 psia.
O i
This test demonstrated that a design pressure of 360 psia is required
- o for the CNSG application since it is conceivable that the Halon piping may be at l
sn initial temperature of 200*F. The NFPA Studard 12A specifies a design l
pressure of 360 psia or 600 psia.
Test 6.
The piping near the di.< charge noszles was modified for this
{
test in order to sinulate a line rupture or a nozzle break. Referring to Figure 4.1, a separate 1/2-inch line was attached to one of the sections of piping that the nozzles are connected to.
This tea connection was closed with a 1/2-inch solenoid valve. During a test the solenoid valve could be opened and liquified Halon would flow uninterrupted from the storage tank through 1/2-inch line and through the I
1/2-inch valve into the chamber. The pressure rise data shown on.no Data Record sheet for this, test show that a vaporization rate of approximately 72 lb/nin was
[
realiced for about a two-stnute period. There was no way to nessure the injection rate during the cpan-line test, but since the pressure did not rise dractically l
after the solenoid valve was closed, the injection rate must have been close to l
56<
T e
4-10 j
l a
i t
4
)
l
4m.
1 (Nl the vsperization rate.
i-d It is concluded f:cm this test that a rupture in a 1301 injection line or vn t a in a real system would not cause a problem in discharging and vaporizing
.,,(
the agent.
l
$R k
ab I
I T4 I
t
?m
{
- w l
i,
i 1
3 I g
. 3 I
57<
l 4-11 I
I I
1 1
c:.
T Ei, 5.0 3EASUPnENT OF ABSOP2 TION EFFECTIVENESS OF CHARCOAL UPON EXPOSURE TO HAiON 1301 g
The possibility has been considered that the charcoal absorption e
beds associated with the contai: ment r.ay be affected in their ability to
)
absorb 1 cud CH I if Halon 1301 gas were present in the atmosphere. The I
2 3
2 1301 could result from either accidental discharge during normal opers*. ion, or from deliberate dischrge followidg a IDCA. In order to determine whether 1301 has any effect on the charcoal absorbtivity, a series of tests were conducted by Environmental Engineering and Testing of Richland, Washington. The ts:sts were performed according to USAEC RDT Standard H-16-1T, "Cas-Phase Adsorbents for Trapping Radioactive Iodine and Iodine compounds," October 1973. The conditions of the tests were as follows:
I Test 1: 25'c 70% Relative Humidity 1 Atm total pressure 20% Halon 130I (balance air)
Test for CH 1 adsorption efficiency 3
Test 2: 25'c 701 R elative Hu:ridity 1 Atm total pressure I
60% Halon 1301 10% Hydtc gen (balance air)
Test for Cd 1 adsorption efficiency 3
I Test 3: Identical conditions as in Test 2, except tedt for 1 adsotition ef ficiency 2
j Ta Test 4: 50'C 951 Relative Humidity 3 Atm total pressure g
60% Halon 1301 (balancs air)
Test for CH 1 adsontion eUiciency 3
I Test 5: Identical conditions of test 4, except test for 7 adsorption efficiency q
The comp 1wte report of the r-J. program cf EET is e.ttached, including J,
their letter of transmittal, stessary sheets of the results, and a description of their test procedure. The results can be stanmarized as follows: For all conditions tested the charcoal that was employed [ North American Carbon I
ss<
l 4
5-1 e
O i y 1
5 s
g
- 1 M
L;1 5
p E,
Tabis 5.1 Perforczaca Requitar.4nts il i
to USAEC RDT Standard M-16-1T fjf Yb E4 Decontamination Ef ficiency
_4 rp I 0 25'c and 70.i RE:
99.0% minimum 2
1 I" *******1#' 130*C, 50 psia:
99.9% minimum t',
2 ha Q116 25'c and 70% R11:
98% minimum 3
I Cl I from steam-air,130*C, 50 psia:
98% minimum h$
3 m
b T.
N.
s
~
I e
U
{
I
']
.q.
NR 1b I
f T
n E
t b
{
eW 4
59<
5-2
)
'=+**-*e
. em. -
m.
O
' y, r
t 6
I j
G-615 5 percent improgr.ated charcoal (KI)] exceedad the Farformance Raquirements (see Tabis 5.1) of the DT Standard (Table I of the Standttd) with the very =4 ar exception that in Test 5 the first 1-inch deep bed of charcoal absorbed 99.82 percent of the I instead of the required 99.9 percent (a 2-inch bed absorbed 2
100.00 percent of the I ).
It should be noted that in control tests at 130*C.
2 I
42.5 psig and 95 percent relative htssidity, in which no 1301 was p' resent, the charcoal absorbed 99.7 percent of the 1 instaad of the RDI Standard requirement 2
of 99.9 percent.
In our opinien, the test results are so favorabia that the possibility of a problem with Halon 1301 interfering with or CH I adt.orption on chard 3
in a containment filter system need not be conaldered.
9a j,
5.1 ENVIRONMDTTAL ENGINEEPlNG & TESTING EIPORT TO ATI.iRTIC RESEARCH CORPORATION OS ABSORPTION EFFECTIVEKESS OF CEARCOAL L70N EXPOSURE TO HALON 1301 (see following pages) rL 1
A
^f ee g
ID 4e dN ap EB e.
a>
.D e
i k
6' T
5-3
.A
~.
l
)
ol l
\\
P. O, BOX 101s MICHLAND. WASHINGTON 04352 May 28, 1975 Dr. E. T. McHale Head, Flacability & Hezard Section Atlantic Rese rch Jorp 9390 Cherokee Ave.
Alexcnl:1r, Virgini. 22314 Derr Dr. 2%Eale:
Enclosed era the complete rceults for the five tests cuthori:cd under your P.O. 1 E-05559. Accompanying these are reshita certifying that the R
Uorth A:nerican Carbon 13o. G-615, used in this test eries, meets the RW-R$
16-lT Stni t
ar-da from t4stin6 utilizing the not a1 air mixttree, preneures and temperatures described in the RM Standard dated Oct.1973.
Also incluied in the package is a generalized description of the teet procedures and technigt:es. Although these are not specific ~ for your test series,.:onsidering gas mixtures, teeperatures and pressuren, they do pm aent the experimental method.
Figure 3 is a current line dravies of the apparatus and in self deceriptive. The lov flow rate of hydrogen gas was measured by a veter displacement methol for accuracy. Testa 4 and 5 vere conducted in acconicace with the procedures described in the Steno-Air section of the RDT Standard 4.5.2 ani 4.5 4; while testa 1-3 j
ccnfor:ed t a 4.5.1 and 4.5.3.
The hydrocen cylinder in still rather full and a co=plete tank of Raion f
1301 remains.
This vos partly due to the mid m in pre-eq.tilibrium time meessary in severa1 teste for the temperature accroso the test Q
bed to attain less than 1 C.
Please indicate.your desired disposition 9
na for these =aterials. We vould he happy to ctore the Halon for future reference cr tests.
i
+
Thank you for this opportunity to serve your Oo=pany.
Sincerely, c
/n Lb A. Vallace hainess Agent
.vm.
9 N
e b
G1<
1 id
=x W
w
. w
. =... - -
.a g;
- i
/
I t
INVIR01HENTAL E!ODGI323G & TESTDD INVOICEf A-1665 DE00NIDMINATION Ef'?ICIENCf 0F ACTIVATED CARBON FOR IDDIRE FORMS CUSTOMERg. Atlantic Research Ccrp.
- URCKASE CRDERt E.oS559 5390 Cherokee Ave.
Alexandria, VA. 22314 CARBON TYFI; c 615 54 Incregnate KI CARBON SOURCE:
riorth American Carben I
CARBON HISTORY:
!!ew I
TEST TEMPERATURE:
25 C
{
TEST FRESSURE:
1 Atmosphere TEST RELATIVE HUMIDITf:
7@
I TE3T FIni RATE:
40ft/e.is
~
1 I
SECTION f CUKfLATIVE EST EED DEPIB C13m.ATIVE DECDlfIAMIKATION ETTICIrr3 10Drtz METEIL IODIr3 1
I f
C-1 1"
99.99 $
i
{'
C-2 2a 100.00 $
99.95 %
C-3 4"
Im,oo g loo,oo g j
C, Backup Beds 100.00 $
100.00 %
1i RIMARK)
Iodine Cc.eentration: 17 5 =;;/m3 Methyl Iodsde ceneentration:
3 I
l 2=a/m I
dan:
5/26 & 5/27/75 f
d' 1
u-blv M
AI. E Ps Ts 7 D."s & Tc.
,4,
\\
P 62<
t 6
0
+.e
,.m
. ao.
.n..-.-e--
.. ~
..-.-.~.-.,,~m
..,,_.J
- l
?y INVIRO!F.E!rfAL E30DIEZROG & w.ma INVOICEf A-1665
)
DEW!rIDMINATION EITICIKICT OF ACTIVATED CARBON PJR IDDIKE FORMS CUS M Atlantic Research Corp.
IG4 M CRDER p E-05550 i
5390 Cherokee Ave.
Alexardria, VA.
22314
.e I
CARB05 TYFI:
G 615 5% Iepregnated KI CARBON SOURCE:
North American Carbon CARBON HISIORY; New TEST TEMPERAIURE:
130 C TEST PfuCSURE:
L2.5 psig TEST REATIVE HUCDITY:
955 TEST FIIN RATE:
ho rt/cin ER0i!CN !
CC}dLATIVE TECT BED DEPIH '
CT_ 4"IVE DETNTAMINATION E."TICIDCY I:n M:lTEYL ICDID3 I
C-1 1"
09.~i"T $
'T C-2 2"
100.00 %
98.6k %
4 c-3 k"
1T M %
99 96 5 C,
Backup Beds 100.00 %
100.00 %
PJ w ava Iriine Concentration: 17.5=.r,Y 5=
3 7
1%thyl Iodide Concentration: F.:=rjn 1
7 2AT3 5/28/75 &5/29/75 g
\\
~
a.
.. L O
~
zr..mac::rrAL Emumns aibrnu v
1 63<
- M
- MM W
% M S.
4@$ gee-M 6M,.
.@pg.
g,,g,,,,
g
-A p
I
=
j l'y i
i ENVIRC10GNTAL ECHIEElRDU & 33TDU DiVOICE f A-1665
~
f DECONIOMHIATION EFFICIENCY OF A:TIVAU:3 CAU!ON FOR IODHi3 FORM 3 a
/
CUSTOMER:
Atlantic Research Cor.;.
KB 3ASE ORDEa f E-05559
^
5390 Cherokee Ave.
1 Alexandria, VA 23314 l
r CARBON TYPE:
G-.619 MT T mrernated &eceal (rr)
CARBON.8OURCE:
North American Carbon I
CARBON HISTORY:
New a
j 0
TEST TDFERAWRE:
25 C
~'
TEST IEESSURE:
1.atmon.
TECT RELATIVE IMCDITY:
70,4
(.
TEST FIIT4 RATE:
40 ft/cin.
/
SECTIONf CU!ULATIVE TEST EED DEFrH 03CLATIVE DE33;fIAMINATION EFFICIC0Y j
!ODETE FEITiYL ICDIC r-C-1 1"
i
~
l~
C.2 2"
99 99 %
i
(,,
C-3 1+"
100.00 $
[;
r C
Backup Beds 5
100.00 %
x i
RE MRKS Three hour pre-elution time. Air mixture: 20% Halon I
3 3
/
1301 (CF Br): 0.175 ft / min Air 1 0.695 ft Alin 3
't
{*
T
.c. q DATE:
5/18/75 i
NN E%, '
k_'
['
2 4
Ci!RCI@"*AL E?CD:EZ'tI D & TZ3Tru G4<
m EO tv
' t 1),s
,g
]
ENVIROWJtEAL EGruERDiG & TETIM l
3701CE ! A-1665 l200N:CMINATION XI'TICIEiCY OF ACI'IVATED CARB3N K3 DDIKI WW3 j
i' C::S'"tNER:
Atlantic Research Corp.
WRCHASE CR::33 Y E-05559 b;
i to Charckee Ave.
Ii ydria,'VA 22514 L
i 4
]
S15 5% 1mpres:nated charcoal (n)
NN TIF1::
CA3333 SO'.GCE:
4h Werican Carbon d
CA3313 EISIDF i
<t 3
TEST 'DEKPERAlt C
]
DIST PU:SSURE:
at=os.
TES" FIAnVE IMh.4 :
70%
]
CST FIG RATE:
40 ft/=in.
n
[l res +
Ca u =v n r azazuem Czum 1.:Cnuu:.no:: crr1CIwt ICDU3 ML"NIYL ICDI2 qb C-1 1"
100.0C%
{i C-2 2"
100,00%
90.91 %
Ll C-3 k"
100. von 100.00 y t
e-
[
q 3ackup 3eds 100.00%
100.00 %
i y y,m Two hour pre-elution ti=e.
Air =1xtures 60$ Halen 1301 3
7 10% Hydregen,3C$ Air Flows rervactively: 0 52 ft /=in,,
I-3 3
0.087 ft / min., 0.26 ft /=in.
- A3:
5/19 & 5/21/75
%g 1
\\'
.w.
a-s dJ fJ i
e
.up mee 65<
s
'THE IODINE COLLECTION EFFICIENCY
{,
i OF ACTIVATED CHARC0AL J. D. Ludwick
]
INTRODUCTION At the present time, considerable effort is being expended to determine the methyl-iodide trapping efficiency of acti-k vated charcoal used in nuclear reactor air confinement systems.
The, concern has been over the possibility of U
a significant fraction of the reactor iodine inventory l
In escaping as methyl iodide in the event of an accident.
the case of the largest water-cooled reactors, further com-j plications could result from the release of gases having j
]
elevated temperatures and humidities.
Under these more severe conditions, it is known that the charcoal trapping
]
. capacity for methyl 1671de is reduced and the degree of reduction is dependent upon the severity of the conditions.
1 Several commercial manufacturers of activated charcoal have recently incorporated special additives into their pro-duction processes in an effort to improve the trapping ability j
C for methyl iodide.
One would expect charcoal to first adscrb methyl iodide and then allow its slow removal by chromato-L graphic elution.
Charcoal additives such as KI, which exchange iodine with the organic material, provide an effective means I'
of removing the radiciodine while treating the radiochemically-m inert methyl-iodide molecule in the usual manner.
l L
A facility was designed and constructed with pressure, 17 l
temperature and humidity control to independently test the
{
ability of these new activated charcoal products EXPERIMENTAL i
i m
1 d
Flow tests were conducted in which molecular iodine was passed through several plastic, metal, and glass tubes to
'. 2 evaluate their reactivity.
Most plastics reacted with or u
adsorbed iodine readily and were not desirable for construc-tion.
Even teflon had a slight affinity for iodine; however,
{*
the adsorbed quantity could be slowly cluted when sufficient 1
air flow was maintained.
Glass was found to be chemically
{
and physically inert to iodine and was used in a limited way.
=
However, glass does not adapt itself easily to an assembly e
66<
I i
N,'
i l
i bl ENm0fMENTAL E9GINERDl0 & TE2CM j
INVOICE / A-1665 i)
DECDNIOKINATION EFFICIENCY OF ACTIVATED CARIC3 ItR IDDIIE FORMS d'
a CUSIOKER:
Atlantic Research Corp.
PURC7. Ass cRDza g E-05559 1
,f 5390 Cherokee Ave.
)1 I
Alexandria, VA.
22314 4'
CARBON T!?E:
0-615 5% Imeregnated Chs.recal (KI) a CARBON 10CRCE:
North American Carbon t-L CARBON EIZDHY:
New j
4 0
TEST TEMF!RAIURE:
50 C 1,
TEST IEESSURE:
3 Atmospheres total l,g TEST ar.ATIVE nutIDm:
95%
f.!
TEST FIA: RATE:
40 fi,/ min total Gas plus Steam.
l SECTION 6 CUKILATIVE TEST BED DEFIH CW.TJC7E DEC0ffIAMII ATION E} TIC!Ci:I Icm MrIEYL IODI:3 C-1 1"
99 825 C-2 2a 100.005 99 86 %
l C-3 4"
100.005 99 98 $
C Backup Beds 1G0.00$
100.00 g x
r 5
To::rporature equilibrium ti=e: 5 tours, steam pressure ag:wum C,
1.6 psia. Gan mixture 60% F.al.~,s 1301, 40% Air i~
CH I concentration:
2=g/m3 3
DATs:
5/22 & sa/75 ce E! 7.?CSCT*IAL E:CU"dE3I:D t..w.;,
Q
- v G7<
_e.:.
.-s,
.ev.
=
-.j
t i
t that must be dismantled without having reactive materials for
{"l connecting joints.
In our experiments, corrosion-tested stain-less steel was found to adscrb iodine slightly.
This surface adsorption was'readily removed by continued elution.
As a consequence of these tests, the facility for charcoal evalua-
]
tion was constructed insinly of stainless steel by using com-pression type fittings (Swagelok) with a limited number of teflon-sealed pipe joints.
Also, some glass was utilized where
[
... appropriate.
Major portions of the test facility are illustrated in b
Figures 1.
and t..
' Air was drawn into the system through a flowmeter (left, Figure 1.).
In tests where humidity control j
was unnecessary, the air may bypass all other adaptations shown in Figure 1 and go directly into the two serially h
arranged tube furnaces.
Yhns adjust the tempera-ture before the air enters the succeeding charcoal test f'
chamber.
J The radioactively-tagged iodine was interjected in.o the air stream either as mothyl iodide from a pressuri:ed stain-less steel cylinder (Figure 1.-) or as chemically-generated
~
molecular iodine from a small glass flask.
The d
injection point was identical in each case.
Incoming air may take one of two alternate paths to the test chan.5er.
A "T" I s arrangement allows this air to be either preheated or hur.idified or both in a stainless steel boiler.
The boiler was constructed out of 5 1/2 in. pipe with welded end plates, and stainless valves on each end allowed I
j complete removal of the boiler from the line.
A sight glass and pressure gauge were attached to the side of the boiler.
j$-
l The sight glass was constructed from thick glass wall tubing and the, ends were stainless steel tube fittings.
The lower I
fitting was connected to a distilled unter source so that,the t
I 68<
-r
,,e-,
e
,v-
i O
(.
'W>
.U a-
.,m
- c..c..;v..w.h a
.. ? A3. " '.n[; r '.A i,
,+ t 0,
.i e
.m,
.o.
.vg.
c t,....
- x. _
.u.m
- J
~
n.
mg,ye. -.a.
. x,. L. 3 4
u 0
., n.. -.)
.g 4 thekf.,..f
- ',45'ic2?b.
?
s Q..%S.W.4.'C.'[.
=&,
- .' S
^ !
1i i
l E.?.y m..
ma
.e w... p ;:
..,v:q.g n..,
3..
.~a:y
.r,c.
.x.
.. :..;,,.. r..., ;;:. M C ;3. g.
,l es
..: = y e
a
..... w'
- r. ;.4 i.
=
u
,s.,
...y..
.c y
e..;.. j o.
z.!
. ;. n....,,
.=.. '
,. s :.. +.,
m.
.. '. c.'..-y, g
4
{
1.
,, c,-
la
/
6 N>.-
' 4.. ' l.,
a ir e
)
. 4,.
k
.s I.
4 l
, f ;:. :-) ?:0... ' q a
a 1,
a
,. b,......
.e,s
.~
V 4.,
m
=o 1
-..i o
s.
+
,n. ;
ou o
s.e
, t,4p,w.., w-P,,. ' "I+s.Al M,.
o m
3-i.6
., u.
n
..i
..p.
s.
- .a e
?' 1, ]...
.a
. x,.'j, o.n.
^
e.
g Au i.
.i.<'
,~'
P, [yY g% Td '..E,..
o 5
a
.i, 'l p k',
f
b I=
. r/q.
- d. y =
...r b.4 a
p
$ ss eeA($d'* "dp-l'
- ' :. e L.
j
=
p d
r a
q g~
y
&. ~g W p'f W,.. e.E..y.>r' =.. e +.= ;.;., -- 1. A [ v,
R s
a p..
i
. M"
,I I
g4 r 8..
s
-w, L.,,
a
'8
',.4
~
s u a&w.
l
{
1 r
l I
(
n 63<
l -
a J
l 1
1 t
i 11auan es f
I" d
w -, -
pr9 t j
. w/ W.* /.,4 t.-
t.+
+
..,n.....
s l.
4......,o.
I,,,,,i,
.s g.fs
-4
'l s
t W' Y!:. 42%)&.ak et ^
- p...
n- ::hq.4.:. Lft.
- Af -W.s Q
[
psen* *..
r+
e i
sw..s.<. c.o V-t.
L
- =; e v
- y.y.*
W..,-e
.c.
2...
%... ;.?. :'
d l
0
{.
'i;A.94l(*g%
^
i-1* *, % Q??. mge,A.
~~~
r0 2Dt
.a e %. (
r.
, w,.;...* :.5 8
I 4
my ) ' e '.
,, ['# "
.b'
~
g/;,
aup -
..,;, @ y 1
-**j
.g
.,6
,T.
9 7..
'.y.-
- q..
i g
e
!.A-y
- j
-;.;,c,.e y f J==*
+,.
5*
f,.
..4., n.', f. *.
+.
R
.e y
. f,.,;, - Sg.;.
L' s),
q g
O %n g b
~
s ;,
f,
',.G.-N.l :.
- J*,,
1.
-.j
,' k
'4 j 8
h e <. w a,
~,..
s a
"y
- +,
e
,= ~ *
,g r.E..
L x
34-
+ *
- g s
a
'(
1,M.,
L J ;m.
,,r.
.e o
+
44
- t
"..' -.M I$
j 0
'e,.*-
.w.
=
lW,..7...),@ @~-} k4.*:y.-
,. R, y, 1" i
~~
e, r..
- x.
n.
- ~,
- w
.n,'. c e.
..[.
h yt t.
1 [;+,
..W 3
-.,, i. )
l w.57/L W,,c ii T
h~.. g$.%a. m?;yr 4
\\ppy/;;.T..
s. p.. w w p,
mA
_n..,, f,,,frt,*), -.
w.l,m. v. =>'
s.
' y',r*,
D p r m
~~
!..01
. pyo%p" dc.,.[ft;;;;,.l% *$b(f %h sm 2 *.
.6 4.m:. upfp. r-n.
m.
c :..
.- :4 0,..:.y gf,
3
[.
f -
. I, 4
,;p!]-
\\*
- v. 'W,ke.)
,...f.:
q,,, p,
.i '
, rs.,y y
p*.
.t u
I.
t.,[ h.,,3
,,, e.
y, y/,[.*
8 h _ff x'.. 'i %L ' '
h%
M M '
- LW.f.r *.
W -
h.
/
9*e()
$ %,1 r
e
m
.42 Am.--
.J
.iA ma
.6 s.tu G-*
4 O
3 9
4 e
8 iinlig P e
9 4
4
\\
r y.as 2b
^ 1 1
gays-.. p SP -
8#*
4
- .',' s.
u.... < :..
' r '.
..'q.~.
I
.,;,9.. >+.:..,+ -
..to.
e a. 7 *,
pa %
(',?.)),,p3';.179.,. :. -
'...e l
..l
.~
..,a.
s
/.
s
\\
- m..:
L *j
's
.g,\\[3.iEd.
e M*
)
+..
J.:rt:n.gg, w
- Q t,
-g:- $+uv:ht'pEr.r.
a-
.fm
- . :.. s..dhv,. str.n g
m s me.w.g.e.: #:~. :
er s.a
. ~ w~
'.j ya s..s w M i
ii$$M~.,.m..g.sni.w..7dd.M..
%d7 i
$. ^:..,
~
[h.,',,$[
- J.
- C.
9 ' ~ * *
. _... ;,s w?...'a>*,,
..e M
.e
..q' h,.i _ 7. e < -,
- u.
,,,9[k.,,,.r.=~ y ),
- f., '..
- I.
{
Wb
.FT-
- 3. '.
., 1,~,",,..
<E,'
. ?,,/-
,;m z
. t.,. r}r;v,.,.
~
^.. -
' ' " %. C k)f. gg(...r :.
'$h Ei r-l'_
)'
- = >'
.s, us.
,4
.,. v )
. ' N,Q]l,,
']
,4,.,
,y n
.s
.v-
,,, w=.,
- ..n,-
e
~
g
,e
.s n.
=.
3: n 3;),s vL.,*; s.
+
n;.6.,m,, e. Q.,,.
2,.p. L
. '. + " ' *p:
v
-Qy.
^
e
.g,.
-s e
~
.,i.
.- vt
~;. :'. ;. \\.,&s. p -
,.,7 i,.
~
^r 6
, 2 *'
4
<e g i
e 1
g
,' ~. 4 R. e-h.
- /T.
5 **,s.R.i r.i e,~ ' - ' J -.
I' o vv e,s,
h *k O 8 44 g **
'.l-y v
8 eSG gg}y ' j._.
p.,a w y,.. a.e.o,e
- - ' h*e=;m, z.
r
. - t. +f,
- o.
- - Q.. tg r-y i
d _.. i 1
3,, -
.r.
g %.
$ M..! @A' A
Ma h
4
\\.
4 9
l 71<
e n
w
,--,*r,--,,,-
... -.,~.,
m,
=
+
i M
M
- - 3 t-,
C3 C,
mW grg G
O O
W C33 Sa8 N
- 1 m
M k
G A"Killi TY Pressure tronomittai J reum Preosuru Contmlor j
Gas l \\Nl i
Mixtures g
Q)
(
compreesor l
Nceiver t
A
)
f DirhrdIntial 1 >
Flovtzster l.-.Y -.V Teet Bede J6 0
g
(
I Heating Zone h
- - *I - - -
l
^
~
A
-T~~~~'
frontalty Ge'ge e
1 '
3 i >
1O P---
---l-t 8
8 Doller
\\
=
e s) /)
/_.____
g T..
e
,o Tube Furnace v
L
'l N r
r t
w
_o A
3 (j
e,ychrometg 7
,i
_l Coulenser e
()
e
/
o
()
_J
'n:sperature I
(
Cont N Uer' Air l'urification Flussster I "3
IodIno y
Fluvuister Raservoti-
'lhennoccul V Flow Control N)
Humidity Renaout i
(diay.hram) g Flow Realout 12 polut
\\
sp Control 3
2
./IgitalControl
[
f I
i Wetor
}___
Inlet Methyl Flow Aljust
.a:_.
j Iodide
-Water Rate
- Pressure Regulatorp Beating Zone h Di[gital
(
)Mcasurement Dehumidifier Final Charecs1 Traps 5
V
/~
~
N I
Control C
\\_/
Flovmeter FIGlEIE 3 UA3 FIIM DIAGUJ4 Vacuum O
Pump l
, -., ~,. _ - -. -.. _ _
m.
m.
4
3, 4
\\'
C, 1
boiler could be filled and the level maintained while in I
operation.
Heating of the boiler was performed by a large three-zone tube furnace.
l Control of the furnace temperature was maintained by a
- s..
O to 600 *C Honeywell controller coupled to a shielded -
1 l
iron-constantan thermocouple.
The controller operated sets of l
heavy duty 110 Vac contacts to the furnace zones (Figure 1 -).
-L t_
h'henever humid conditions were required, the air-iodine gas 1
was always heated further in the two tube furnaces.
This j
{-
prevented condensation from collecting in the stainless lines.
1 These two furnaces were controlled in the same manner as the 3,.
boiler although the controller unit is not shown in the figures.
i i.
The air-iodine mixture then entered the test chanber whero it i
was thoroughly mixed and evenly distributed by a perforated t
<l
(,
deflector in,corporated into one end of the chamber housing j
(Figures 2 a and 2 b).
The gases subsequently passed into a f
two-ecmpartment section with a thermocouple fitting shown C
directly above the perforated unit, y
m The test chamber, 2 in. ID, housed two 1 in. deep beds of
(
charcoal separated by a thin perforated stainless screen.
(
j These charcoal beds were held in positics by a stainless
{
steel sleeve which was drilled to allow the passage of a
,[
thermocouple to a position between the upper and lower beds, i
Two additional chromel-alumel thermocouples were pinced at the 1
i a
1 4
1 4
'73 <
4,
l O
l; i
n U.
l
~
entrance and exit positions of the chamber'.
Ordinarily, only i
}
cne chamber section containing a fi'nal 2 in, oeep bed of i
activated charcoal followed the two-compartment section but a series of 2 in. deep beds could be added to the chamb'er by inserting additional chamber sections as just described.
The sections separated by perforated screene were internally
{'.
i sealed by the use of 0-rings mounted in grooves in the chamber j,
section walls.
Pressure'was applied to the 0-rings to insure L
a tight seal by three threaded rods holding the sections together.
All threaded connections were teflon-taped for I..
sealing purposes.
The stainless steel sheathed chromel-alumel thermocouples were connectel to an 8-point recorder whose I
j temperature range was adjusted with resistor boards to the designed range for a test series. The thermocouple readings 4
i
'1 I
were recorded in sequence at 5 see intervals.
The gases in tests conducted without additional humidity
{
adjustment were exhausted at the exit of the test chamber.
Hi,gh humidity experiments utilizing the boiler. required the j
remainder of equipment shown in Figure 1.
The air stream was i:
diverted thr'ough the. SS condenser coils, which wore operated in parallel, to allow the high flow rate desired.
Condensed water from this air was also forced up the coil to the over-flow bulb located at the top of the apparatus.
Here the l
j air-water mixture was separated and the water was channeled to f
the icwer SS reservoir.
Rather large amounts of' water (i
were collected during some tests, and up to 2 5 liters could h
be easily handled.
The reservoir was internally calibrated
{"
for water volume, and each test could then be calibrated for
)
humidity purposes.
A Valve allowing rapid water removal jl was attached at the bottom of the reservoir.
For.high-
}
j' humidity long-term tests, the water in the boiler could be y
L
{
I.r
\\
J 74<
. =.-
i replenished and the reservoir exhausted at calibrated. intervals,
{!
thus allowing continuous operation at high humidity.
The cooled moist air was then led to a small tube furnace-ji where it was warmed to reduce its relative humidity and was subsequently passed through a 3 in bed of activated chercoal contained in.a stainless steel sampler.
L The purpose of this final trap was'to ensure the complete 131 removal of all iodine used in the tests.
The absence of I
..,j activity in this trap throughout the test series provided proof of the effec _tiveness of the test chambers' charcoal beds.
)1 The gas flow in the entire system was maintained by pump or g
vacuum supplied at the exhaust end of the final trap.
This
[
l connection was not illustrated.
l I3I PREPARATION OF N0LECULAR 1 TAGGED WITH I
2 1.
Generally, the charcoal was tested in the facility with a loading of about 100 ug molecular iodine per gram of charcoal
)'
in the upstream bed.
Figure 3 illustrates the iodine generation
~
position and the flowmeter through which air was drawn into j
the overall, system. Air flow adjustments were made by down-stream valve control so that the exact linear gas velocity in i
the charcoal bed could be maintained. Molecular iodine was y
u prepared by the following reaction:
[
'i 21" + 4H' + 2N0 ' + 2NO + 2H O + 1 2
2 2
A measured quantity of sodium iodide was added to a 3N H SO 2
4 11 solution containing sufficient acid to complete the reaction 131 as written.
To this was added a quantity of I in basic sulfite solution sufficient to tag the inert iodine and allow
' i easy analysis of the sampler sections. About 5x 10~3 uCi of i
131 the I were introduced as a tracer in the test and the iodide
(
i l,
e a
l
[
75<
i
~ ~
~ ~ ~
~~
1
, }?
was oxidized to iodine by slow addition of nan 0.
All iodine-2 y
and held containing solutions were treated with Na 303 2
y slightly basic to keep the, iodine in the iodide form prior to The rate of iodine addition to the air streca (and the d
]
use.
J resultant iodine-air concentration)'was partly centro 11ed by j
the NANO addition. The iodine formed in the generator was 2
swept from the solution into the air stream by using helium This maintained positive pressure in the generator which gas.
g prevented air entering the system from bypassing the flowmeter.
131 PREPARATION OF METHYL 10010E TAGGED WITH g
~
Efficiency tests of activated charcoal required the prep-aration of methyl iodide having a known and useful amount of L f 1
I tracer.
An experis. ental apparatus was set up for the preparation of the iodide which consisted of a three-neck flask n]
connected to a water-cooled condenser (Figure 4).
A thermom-
.eter w.s placed in one neck and a separatory funnel in the j
remaining neck.
The flask was wrapped with heating tape which was.. temperature controlled by a variable transtormer.
A mix-131
[
ture of 8 g KI, premixed with about 0.01 mci I in the form of a NaI solution; 4.3 ml H 0; and 0.6 g CACO 3 powder was 2
"j, placed into the flask containing a magnet stirrer.
The flask i
was ware-d to 65 'C while stirred slowly.
Drops,of dimethyl sulfate (4.7 ml) were added,to this mixture from the attached separatory funnel.
The product, formed by the reaction, 4 + 2CH I* + K SO4 f
2 KI * +. (CH ) 2SO 3
2 3
was slowly distilled from the mixture and condensed into a at 0 'C.
The dry collecting vial containing anhydrous CaC13 methyl iodide (4 ml) was decanted into a storage vial containing a drop of mercury to prevent iodine formation.
The vial was stored in a deep freezer, and aliquots of methyl. iodide were taken by using preccoled pipettes.
4
=
76< esme=**=m e-
- 4
.ame 4-w....
- 8""*'M O th' a
..g
C C C.- C CD C; E (. E
- - C'?
E i i .i Separrayfunnel with (CH 150 [ 32 4
== t I Condensor (' Thermometer {<,h M
- ]
i ^ \\ l N .w l l ~ 8 .. v Collecting Vial ~ w Heating Tape Cold / with CaCt Bath" 2 "j' KI + %0 + CACO + I3I l 3 s i Magnetic Stirrer l 1 FIGURE E. Apparatue for Preparation of Tagged Nethyl Iodide. I i L~~'~~~ ..,. f.'.'I,1.'C.~ *l*.'.I. ~ * ' * * * * " '"W 6_
~ I. 7 + ~
- 3 3
Figure 1 illustrates the one-liter stainless steel Ir cylinder with sensitive pressure gauge used for dispersing the l, . radioactive methyl iodide. After the cylinder was precooled in the deep freeze, 100 911ter aliquots were transferred to the
- I open cylinder. The cylinder valve was replaced, and the out-let was connected to a high-pressure argon gas source.
The i i' cylinder was pressurized to 100 psi and then allowed to stand overnight to facilitate mixing of the argon and methyl iodide. 1 The vapor pressure of methyl iodide at room temperature was ) sufficiently high to ensure complete vaporization of the com-pound within the cylinder, with the argon gas acting as a carrier for iodido dispersal. By carefully bleeding the cylinder into the flow stream, the resultant air-iodide con-centration was controlled. Air concentrations of methyl 3 iodide were typically ' 2 mg/m, and the charcoal loading for methyl iodide'was about : o.h ag/ gram charcoal tested. ~ CHARCOAL TESTING The test chamber was assembled as follows: e About 3Q g of chcrcoal were placed in the first 1 in. bed of the primary test section. The material was agitated to provide close packing. The separator screen and steel sleeve were inserted ~ and aligned to allow the center thermocouple to be woanted at the time. Thirty six grams of.charcaal were also placed in the { second bed.
- The final two inch chamber section used for backup purposes was similarly prepared.
The sections were assembled to form the test chamber. e Three bolts that produce the pressure seal on the 0-rings in each section were tightened firmly but without excessive pressure. l l i 7S< !f i i
1 ] e A vacuum was drawn from the exhaust end of the system and 1 the test began. No other control was necessary for charcoal tests at ambient temperatures. Generally, the iodine was loaded onto the charcoal in about I 30 min. Air elution continued fc1 2 hr after test initiation ,~ in a typical experiment. When the iodine collection et'ficiency I of the charcoal bed was measured at elevated temperatures, slight changes in operatzng procedures were used. The air j ficw was adjusted to the proper level, and the two-series tube L furnaces were turned on. The temperatures of the air.in the charcoal beds-were measured by the three chromel-alumel thermo-I couples in the test facility and recorded on the 8-point ? recorder. Furnace temperature control was maintained by using an iron-constantan thermocouple and a relay operated controller. When the air temperature reached the desired level, iodine was + l> introduced and the experiment pre eeded. Temperature control ~ was generally within 1 *C of the desired temperature, with i smal'1 adjustments made in the controller-thermostat level. I To test the capability of activated charcoal under rather extreme operating conditions, taxperiments were con-i ducted under humid conditions at elevated temperatures. The furnace used to supply heat to the boiler was turned on with the boiler sealed from the system. Air flow through the ~ system was started and the tube furnaces turned on; all con- } trolling and recording systems were operating. The pre-exhaust heater was turned on and :et at a point' which provided about 50 *C air to the ed.aust trap. This greatly reduced the air humidity prior to passing through the final charcoal trap. l As the water in the boiler was heated, the internal pressure ~ rose and when it reached 30 psi, the downstream boiler valve was slowly opened and moisture entered the air stream. Approximately 1 hr was required to bring the boiler to 1 79< I L
- {
- l b,
temperature. The upstrata valve was then opened and the } l3 f j bypass valve closed, fcrcing all incoming air from the flow-meter to pass through the boiler and to become saturated. j Heating and humidification were continued for at least anothar a hour while the temperature and humidity in the charcoal beds I 4 , stabilized. During this time, the cond'ensers were in opera-j tion collecting water. When a satisfactory stable condition I was,noted, the water level in the collection reservoir was j measured, the iodine generation was initiated, and the l recorder was marked for operation. Adjustments in tube and j 1 boiler temperature were made to maintain the desired conditions. .j u Practice in this technique was necessary to beconej amiliar i: { with the particular characteristics of the system. The rate of water condensation was in.iicative of the humidity of the i operation, and it was relatively easy to maintain almost L 100% humid conditions in the test chamber. The average [- humidity during a test was ca.lculated from the quantity of L water collected end total air flow through the system. Cor-rections were made for the water remaining in the exhaust air at the exhaust ges temperature. The quantity of water collected was in agreement with the theoretical anount expected in saturated air at 100 *C. Air saturated at 100 'C l; 3 1 contains about 600g H 0/m ; therefore, quite a large volume 2 f of water was collected in each test. At the completion of a test, the height of the water in the reservoir was noted. f After the facility was shut down and air flow was terminated, the test chamber was disassembled, and the char-j coal from each bed was transferred to a polystyrene screwt'op jar which fit into a well-type ganza-ray spectrometer i '(9 in. NaI (T1). Similar measurements were also made on the I charcoal in the exhaust' trap. The total iodine collected on 131 each bed was calculated frem the amount of I gamma ray f [ detected,underthe0.35HeVphotopeak. Samples of all I. 5 E SO< 2
1 I ll charcoal types tested were taken for T-ray analysis to measure 131 the possible interference in the I pasa energy region. These measurements showed no unusual background interference 4 1 i level.s for new charcoal samples; however, several radio-I, contaminants were found in used, reactor charcoals. The amount f 131 of this residual interference found in the region of the 1 gamma spectrum was used to determine the quantity of tracer j actilrity to be used in the experiments. In this way, inter-l l pretation of the results was simplified. u 'J 1 Imd .M J Ml !J q J 'l i.; I i 1 k t t 1 [ Si< J
-i 1 mu i i u ij 6.0 RAD 101,YSIS OF HALON 1301 't q J h. 6.1 DISCUSSION OF KEY QUESTIONS t If Halon 1301 were injected into a containment it would be sub-h jected to a radiation field which could conceivably lead to decomposition. f In order to determine the extent of radiolytic decomposition to be expected, an analytical study was performed by ifuS Corporation. No experimental work vas,to be conducted, but required information was to be obtained from the technical litarature to the extent possible. The full report submitted by NUS is included at the end of tais section. j The conclusion of the 10S study was that the direct radiolytic I decomposition of gaseous 1301 in the containment would be negligibly small. J An estimated 0.00023 grame per day per R/ hour would be the decomposition rate. However, two other questions were raised by NCS in their study. In the first, they identified a possible alterr.ative machanism for the decompo- <i sition to occur following a IDCA. This alternative mechanism involves the dissolution of 1X.1 in the cooling water with subsequent decoc: position of the 1301 r i ha cooling I t water passes through the reactor core. The postulated sequ=~. r' events a is as follows: A 5-inch surge line is assumed to bred. (dc . t seis d accident). Eventually water at a race of 500 sps woul. he ;mp. through this line, would fall into the containment sump, and t uu 9e .iteulated f I continucualy via the low pressure injection system throup tha Mactor core i ~ and out again through the break in the 5-inch line. As th.etar passes f U] { through the containment vessel atoosphere, it could dissoln, wt quantity I of Halon 1301. As it next passes through the reactor ' core. : N 1301 could decompose indirectiv by the reactica 3+E[ g ~ CF Br + e - CF 3 10 l which may be extremely rapid with an estimated rate conscent of 2.5 x 10 liter / mole /sec. Even though the concentrations of the reactanta, CF Br and l 3 l hydrated electron, are relatively lov, it is still possible to calculate a large amount of decomposition of the 1301 over a 67-day period following i a IDCA. k-82< 6-1
The amount of decomposition that one calculates depends on what i is assumed about the amount of 1301 that diss61vsa in the atar..If the rate of dissolution is assumed to be infinitely fast; i.e., all the water at all times contains the saturation amount of 1301, then it can be shown I that virtually all the 1301 would decompose in 67 days, if the reaction above is as rapid as assumed. On the other hand, we at Atlantic Research a have parformed calculations based on a dif fusion-limiting model for the dis-solution process which in4cate that lors than 1 percent of the 1301 might dissolve in the H O over a 67-day period. Accordingly, if the model and 2 calculations are valid, then the decompocition of 1301 would be insignifi- ~~ cant, even if the above reaction is rapid. .) It therefore was concluded that an experimental study should be conducted to settle the above question. The study may eventually have to P '~ consider both aspects of the probics; name'.y, the rate of solubility of 1301 into water, and the radiolytic decomposition of the Halon in water. r'I l solution. However, either of these processes could be rate-limiting. I Accordingly, it is logical to first study only oce of them. If the onc g., selected is in fact rate-limiting, then it will be unnecessary to investi-gate the second. On the other hand, if it is found that the process selected l could lead to appreciable decocposition, then the other would have to be studied, The process selected to be studied initially was the solubility rate of 1301 into H 0 This was chosen because it ir the first step in the 2 two-event process that would load to 1301 decomposition, and more importantly, it is in our opinion the step that is most likely to be rate-limiting. The 1 rcoults of this study are prescoted in another section of this report. The second question raised by NUS in their repor. (p. 7) concerne in the e ntainment atm sphere to pro-a possible reaction of CF Br with 12 3 duce CF 1. There are two approaches that can be taken in censidering this 3 { question before having to perform an experimental study of the reaction. One approach is to test whether charcoal would absorb CF 1 and BrI, the 3 i two products of any rarction. If this were the case, then no problem would exist. ':he second approach is to perform a chemical kinetic analysis of i the reaction. Such an analysis was performed for the homogeneous gas-phase reaction, and this analysis is presented in the following part of this section. ( 83< i-6-2 L
m I 4 i-p h 6.1.1' Reaction of I withCFg l g. The question has been raised by NUS Corporation as to whether Balon 1301 (CF Er) might react with 12 in the containment during the 67-day [ 3 residence period to produce CF 1 and Br!. This is a point which deserves to 5 3 be considered and we have performed some thermodynamic and chemical kinetic j computations in an attempt to determine whether this chemical rasetion could ocent to a significant extent in the gas phase. }
- 1
(, First we avamir.e the thermodynamic aspects to ascertain how much '{ gl CF I can theoretically form. Standsrd JANNAF thennochemical data are used 3 throughout. Fer the reaction CF I + BrI (t.H = +9.2 Kcal/mele) (A) CF Br + I + 3 2 3 ~0 a value of the equilibrium constant, Kp, at 323' K (120' F) of 2.00 X 10 L i .f is calculated. Assuming 693 grams of I2 (source of information: NUS) rw-leased into the 34,000 cu fc containment, and 15,000 lbs of 1301 present, the arount of I2 that could be converted is compu:ed to be 127 grsas, or 18.37. of 1 i,i the total quantity present. 8.J It next reseins to determine to what extent reactiou (A) cau actually l cec:- by axamindog the chemical kinetics involvd. The first reaction step l to u nine is the direct reaction as represented by equatics (A). Th2 rate of this reaction is given by i = k [Ia}[CF Brj (a) g 3 In order to compute the rate of reaction of 1, i.e., -4(I ]/dt, a value of 2 2 kA must be estimated. The Arrhenius enwssion for kA is AA exp (E /RT). j A A standard method of estir.ating the activatica energy EA is.to use 28* of the i sum of the energies of the bonds being broken and add the endothermicity of the reaction (+ 9.2 Keet/ mole). The C-Br bond dissociation energy is 69.4 Kcal and that of 12 is 36.1 Kcal. Therefore, an Ea of 39 Ect1/toie is anti-mated. (Reaction (A) is A four-center rolecular reaction, which type 10 very f unconsnon gd vill have a high activation energy.) A pre-exponential factor, L AA, cf 10 cc/ mole see is assumed (a conservatively high value). Inserting these values and the aforementioned concentrations for 12 ) and CF Br into. equation (s) one obtains 3 1.1 X 10~7 grams /67-dav/ tot 1 co.tsi sent (b) [ = Clearly this is negli-ible reaction rate, which maans that direct reaction via l ( l i 1 i 8'N-6-3
i ,3 1 l i # equation (A) will not contribute significantly to 12 reaction. Iiowever, other {, reacticn routes nay be possible a:xi abould also be ext. mined. l sJ The naxt logical route to consider involves the reaction of I2 with initial reaction products of '7 3r, namely, l 3 + CF I + I (1H --18 ) (2) CF3+I2 3 L Br + 1 + BrI + I (18 = 6 ) (3) 2 t Since these are both exothermic radfrm1 reactions, we will simply make the U assumption that they can occur with no significant steric or activation energy barrier. In this case then the aucunt of 12 rascisd will be equivalent to the amount of CF Br dec.omposed radiolytically. (This would represent what NUS I 3 ,J refens to as the radiation catalyze 4 reaction.) Since it ie estimated (see EUS report) that J.000223 g/ day /1/hr of 12')1 v111 decompose in the coatainment. and since sach mole of 1301 would lead to the decomposition of 2 moles af I. 2 then it can be calculated ti.at 0.051 g 1 /67-day / tots 1 contaicment/R/hr would 2 react. This represents 0 74 percent of the total I2 at a dosage rate of 100 R/hr. Because of the assumptions of no limitations on this reaction this will represent a marimum enount of conversion by this route. i There is anothtr imagir.ahle mechanian for the reaction of 1301 with g-13 which involves a free-radical chain. This mechanism is sow.ewhat more com-i plicated to analyre kinetically. Fmaver, because of the significance of the problem,we have performed the a.nalysis and present it for consideration. There are innu nerable ele::.enta y reactiens that one can write when censidering a a cemv11cated chemical process. 1;e have listed below what wn consider to be the L important steps t5at could be oce.:rring in a chain reaction of CThBr with 1. 2 CF 3r CF P Br initiation (1) 3 3 I+I (AH --18 Kcal/ mole) propagatica (2) CF3+I2 3 ~~ i Br + 1 - Br! + I (1H =-6 Kcal/mola) propaga*. ion (3) 2
- r L
- v. + Cr nr cr I + Br (an = + 15.5 veal / mole) propagation (4) 3 3
I+1+M -I.+M (in = -36 Kcal/ mole) termination (5) l As can be seen, this r..schanism ec :verts the prev'.ously-written reactions (2) and (3) into a clai. scheme by e.ociling them with the 3-center displacement reaction (4) to previde chain prcr gation. s Die rute of consitmptien of 12 will be given as h 2
- h 1.
65< l. 6-4 b
p, n h J 11j' m
- l A standard kinetic analysis can be applied.to thin mechanism by
{ . making the (good) assumption that intermediate species are in steady state. Bance, for example, = k [CF ][I ] k [Br][I I ~ # 3 *lIII ~ 5II I 3 ~ 2 3 2 3 2 4 i Writing similar expressiops for [CF ] and (Br], one can then.olve the result-3 p } ing eqcations for - d[I2]/dt in tenu of rate con:tants and concentratiens of stable reactants. Thu4, i 1 d 2K [U Rr[ ,g_ g 3 () de k I 5 ], H where Kg is equal to - d[CF Br]/4c, taken to be 0.000223 g/ day /R/hr. and 'H] .1 ~ 3 is total gas concentration which.can be sat aqual to 2[CF Brj as a good 3 approximation. Equation (e) therefore reduces to i' '~ i .l_ j ~ ~g K kf[CFBr] 3 2 de k5 q The following numerical values can be cssigned to the quantities of j equation (f)r 0.0223 h/ day / containment /100 Rs'hr K = q y 1.8 X 10'10 moles /cc/sec/100 R/hr = 3 ) 'J 14 1-10 exp (- 20,000/RI) k = j g 1 10 cc/mola/see at 323' K ll = i 6 10 cc / mole 2,,, 2 i2 k = 5 []. -5 5 z 10 noles/sec ll [CF Br] = 3 ~ n e activatig ene*gy of 20,000 cal assigned to kg and the pr.4-exponantial 4 factor of 10 grbothcopservativevaluesfortMat.se of reactisn, as is the value of 30 cc'/nole' sac for the race constant.i reactior (5). In-serting these values into equation (f) one obtains d s-l $ l 86< g T ll 3 6-2
I t l - d(1 I 2 -19 9.5 % 10 moles /cc/sec/100 R/hr (g) de - d[I }y 1.4 grams /67 days / containment /100 R/hr (h) dc = i i The rate given in equation (h) is less than that computed for the non-chain reaction (5.1 g/67 days / total containment /100 R/hr). This corres- '[ ponds to a chain length of less than unity. The physical meaning here is 2 vith CF Br, if it occurs, will be primarily by a that the reaction of 1 3 non-chain reaction under the conditions specified. 4 f In addition to the above reactions we have also considered the o. reactions of 12 and I with hydrogen, oxygen and water vapor. We are unable L t to identify a mechanism whereby any of these species could interact to any h appreciable extent. At this point we conclude that the maximum possible rate that the reaction of 12 with CF Br in the gas phase can occur could result in the con-3 version of approximately 5 g of I2 in the' containment in 67 days at a radia-tion dosage of 100 R/hr. L. This maximum possible rate assuces that the species CF3 and Br from CF Br radiative decomposition only react with I. However, in the absence of 3 2 r-12t.these species react as follows:
- I CF Br '
- CF3 + Br (1) 3 Br (6) CF3 + CF Br - CF4 + CF2 3 -CF I} i' CF3 + CF3 26 1. CF Br + Br CF B# (0) 2 2 2 Br + Br + M 3r2+N (9) l-These reactions account fairly well for the product distribution when CF Br is irradiated, namely, 3 CF Br CF + CF + F # + Br 3 4 26 2 2 2 2.1 0.62 0.24 0.98 0.22 I the numbers representing G-values. Reaction (6) ce=petes with reaction (2).for CF. The relative rates 3 of these reactions can be estimated: 87< l 6-6 i
. ~. -....... -.. ii i 1 L f j-Rate (6) 3 I 3r] A exp (-2 !RI} 6 6 Rate (2) (1 } A exp (-E /RT) 2 2 2 1. The ratio (CF Br]/(1 ] is calculated to be 16,742. We vil.1 assume that A6* 3 2 A. Values for the activation energies of E6 = 6 Kcal and E2 = 2 Kcal (i.e., 2 ' g [' approximately 5.5% of the energies of the bonda being broken) can be taken. inserting these data fnto equation (1): 1 Rate (6') 33 k Rate (2) 'l This means that the aforementioned marimum possible rate of I2 de-c. composition (5 g/67 days) should be reduced by about a factor of 2. Further-i more, competition for Br atens will further reduce the rate, although it is not possible to estimate the factor as was done for CF. The resulting maxi-3 . [' mum rate will be approxisately 2 g I2/67 days, which corresponds ta a conver-l sion of 0.3 percent of the I
- 2 b
6.1.2 Conclusions from Analysis of I., Reactions i The foregoing analysis considered the thermodynamics and kinetics ca the reaction CF Br + q - CF 1 + Br1 W 3 3 l The conclusion from the analysis is that about 2 grams of 1 e uld react 2 in the gas phase with 1301 during the 67-day period in the cectainment, l regardless of whether the reaction is radiation cataly:ed. This repre-i sents 0.37. of the I released into the containment. It is. conceivable 2 j' that certain surface areeswithin the containment could act as catalyst to increase the extent of reaction above what was computed for the gas a phase. A heterogeneous catalytic reaction on an unspecified surface does not lend itself to analysis as does a gas reaction. I 4. 9 ss< 4-7 l' l
. - - ~ _. -.... ~. .. ~ _. I I
- i 1 l 1
i 6.2 NUS CORPORATION REPORT TO ATLANTIC RESEARCH CORFORATION CN
- BADIOLYSIS OF BROH0TRIFLUORomEANE (HALON 1301) IN A i
. COMBUSTIBLE GAS CONTROL SYSTEM FOR THE MARITLE REACTOR" a ed t B a i i I f 6 4 4 I 4 he b 4 0 89< l s-a r
,._....--..-..,:--.-.---.~..-.---.--.- . _ _ _ rma ca. weenw,vw.- t [ NUS.1406 l r I i RADIOLYSIS OF BROMOTRIFLUORCMETHANE (HALON.1301) IN A COMBUSTIBLE GAS CONTROL SYSTEM FOR THE f I t, MARITIME REACTOR .ll l.u 1 1 t I l Prepared For The j ATLANTIC RESEARCH CORPORATION ALEXANDRIA, VIRGINIA April I,1975 By i S. E. Tu rner, Ph. D. i Southern Nucteer Department [^ NUS Corporation 1. 2536 Countryside Bouleverd l Clearwater, Floride 1 I e 90< .)
l i-.. . - - ~ g-t 4 1 TABLE OF CONTENTS l. INTRODUCTIN
- ih 11.
SUMMARY
, CONCLUSIONS AND RECOMMENDATIONS A. Summary B. Conclusions C. Recommendations lit. LOSS OF COOLANT ACCIDENT SEQUENCE IV. RADIOLYSIS OF HALON. AIR MIXTURES .= ^ V. RADIOLYSIS OF HALON IN SOLUTION A. Halon Solubility B. Radiolysis of Water C. Halon Solution Radioly h VI. EXCHANGE REACTIONS REFERENCES APPENDlX A-POST.LOCA HYDROGEN CONCENTRATION (SNEH2 Code) g M f* %.T j We 31< ~ + ,n
~- q i 1 L, l l l l
- l..
i fI LIST CF FIGURES \\ I i 1 i j Figure 1 Halon Decomposition by Radiolysis for en assumed . { G (.CF Br,' of 0.5. 3
- 6..
e Figure 2 Schematic of Normal and Emergency Cooling Systems [ ij Figure 3A Accumulated Hydrogen Concentrition in Containment Atmosphere (Radiolysis only. Reg. Guide 1.7 (Version 2) Guidelines) a 5 f Figure 3B Accumulated Hydrogen Concentration in Containment Atmosphere (Radiolysis only, Reg. Guide 1.7 (Version 2) e i Guidelines) i Figure 4 Time Dependent Rate of Halon Consumption Is. i 1.' f r ...l. 9* I f i f I' 92<
= --m.,. = ~ IiU l RADIOLYSIS OF BROMOTRIFLUOROMETHANE (HALON 1301) IN A j COMBUSTIBLE GAS CONTRCL SYSTEM FOR THE MARITIME REACTOR k 'l 1. INTROO*JCTION } 6 8 7 3' The study reported here was undertaken under subcontract to the Atlantic Research !L Corporation, Alexandria, Virginia, for the purpose of investigating radio.We decompo-l sition of Halon-1301. Halon 1301, also known chemically as Bromotrifluoromethane, has been suggested as a chemicaLinerting agent that could be injected into the contaire I ment atmosphere following a postulated loss of coolant accident (LOCA) to inhibit j ..] burning of any hydrogen that might accumulate. For this application, the decompo. sition of l'alon 1301 by radWys.s is a potential problem requiring evaluation. It is this radWytic decomposition that is assessed here. In addition, the cortsequences of irk l advertent operation of the Halca injection system during normal reactor' operation are l also red. Throughout this repor*, the term "Halon"is used to refer to the specific chernical Halon-1301 or Bromotrifivoromethane. n t. I. l. 9 i i l 93< i
r - ' ~ ' ' i.!L 11.
SUMMARY
.. CONCLUSIONS AND RECOMMENDATIONS i +- d 1, A. Summary j' l A literature sutvey revealed very little experimental evidence directly related n. to radiol / tic decompositjon of Halon under the conditions expected to exist within the containment following a postulated LOCA. In one set of experi. ments,1 pure Bromotrifluoromethane was irradiated in the liould state. A total G (.CF Br)# factor of approximately 2.1 was observed, yielding a mixture of 3 i~ j tetrafluoromethane (G = 0.62), hexafluoroethane (G = 0.24), dibromodifluoro-i methane (G = 0.98) and elemental bromine (G = 0.22). It seems likely that a g L-small amount ctf fluorine may also have been formed but r.ot reported. The t direct radioiysis of pure Halon, however, may not be of particular significance i L. in assessing decompositio'n under post LOCA conditions, for the following reasons: (1) in the gaseous state, oxygen, hydrogen, and water vapor would {, also be present and could ' alter the overall radiolysis products, and (2) in aqueous solution, the very limited amount of Halon dissolved would result in the absorption of only ind;nificant quantities of the radiation directly by I' Halon. Chemical reaction of Halon in solution with transient highly-reactive free radicals formed from water radiolysis is suggested as 'he most likely mechanism by which Halon could possibly be decomposed, although conclusive evidence 2 in which is not avaitatie. The literature survey yielded only one investigation aqueous solutions of Halon were irradiated. Unfortunately,in this investiga-1 tion, the irradiation of aqueous Halon solution was employed only incidental-l l' ly as a source of the CF3 radical in order to observe its reaction rates with I'
- vari.,us chemkal additives introduced into the solution.
Consequently, although ro direct experimental evidence on radiolysis of Halon in solution has i i been found. the information given in Ref. 2 is useful in estimating a possible upper limit for Halon decomposition by the mechanism postulated here. i rne G vaiue tofers to the net observabie yieie in moiecuies per 100 ev of eeergy ab,o, bed. 4 I" i 4 l j h. 34< t-l
p. ~,
- l
~~ C. Ii ~ 6 l For a postulated LOCA, us:ng the assumptions specified in Regulatory Guide 1.7, Mod 2 (Draft, August 1$74), hydrogen from the radiolysis of water could accumulate in the ccmtainmentatmosphere in an amount that could require irk t. Jaction of Halon approximately 6.6 hours after the LOCA. Direct radiotytt decomposition of Halon would be neg!igible, part!cularly in comparison with the potential decomposition that could result from chemical reaction in soks-tion with intermadh products of water radiolysis, principally the hydrated-electron free radical Although no experime.tal evidence has been found on I~' i the radiolytic h, ition of Halon in solution, or on the products of such radiolysis, an appromirnate upper limit may be established by assuming that the Halon competes sucrossfully for all of the hydrated electrons produced and that no recombination occurs. For these assumed conditions, Halon in solution I, could decompose weth a maximum G value of 2.65 to 2.7, equal to the vaaue i for formation of the frydrated electron free radicat. As Br* accumulates in the solution, however, the strong affinity of Br* for OH radicals would compete with CF radical decomposition reactions, and result in significant recombuse-3 tion of CF Gr. For this condition, a G ( CF Br) of 0.5 or less would be mcse 3 3 credible. l It should be noted tnat the suggested model for CF 8r radiolysis, 3 i CF Etr + **,q _ CF3 + Br* k = 2.5 x 1010 y-1,c.2 (1) t 3 3 yields the CF radu, which will undergo further reaction. it is possible that 3 CF3 decomposes threugh various rapidly.hydrolyzed intermediates to yie6d (* carbonate &nd the F~ and Br* ions, l.e., 2. (2) CF3 + CH
- {CF 0H]
- CO * + 5 H + + 3 F* + Br* 3 3 if hydrolysis is compoete, the end restalt will be a yield of HF and HGr with G f, values of C.1 and 2.7 r6spectively, Howcar, the Br* ion cornpetes very effk ciently for reaction wnh OH and ter: tion (2) abova will likely not be corap;<+e. j particularly as the romide ion concentration in solution gradually incremes. This could *esWt in s@ificantly smaller values of G for HF and HBr, ee 95<
pp y-
= I J l 1 ( B. Conclusions 'I* 1. Radiolysis of Halen in Solution l Under LOCA cor'ditions, Halon is estimated to be soluth in water to the 1 extent of approximately 0.08 grams per liter, based on the reported solu-bility at 770 F and 1 atmosphere pressure.3 The total quantity of Halon in solution shirtly after injection into the containment is estimated to be a maximum of approximately 12.3 pounds. This value assumes injection ,l of the entire inventory of Halon that might be needed later in the accident l l sequence when H2 might possibly accur. ulate to higher concentrations. No credit has been taken for partial Halon injection early in the accident sequence, with subsequent addition of more Halon as needed. f f*I-- For a G (CF 3r) of 2.7, the rate of aqueous Halon decomposition 6.6 3 l; hours following the LOCA (maximum instantaneous rate) would be suffi- ,l ( cient to rapidly decompose the small quantity of dissol'ved Halon. This potential rate of Halon decomposition, however, would decrease with [. time as the fission product activity decays. The maximum instantaneous l Halon decomposition rate clearly represents an extremely conservative upper limit (equivalent to S.4 moles of Halen decomposed in solution for ea:h mole of hydrogen produced) and further essumes that Halon in solu-tion would be continuously replenished by resolution from the contain, ment atmo phere. Clearly, if Halen radiolysis proceeds at the maximum possible rate, the amount in solution would be consumed in a relatively short time, al-though dissolution of Halon in the atmosphere would tend to replenish that dep!eted. If one assumes operation of the low pressure coolant in-l ! jection system in the recirculation mode (taking suction from the drywell I sump with the recirculttion flow spilling from the broken coolant line), i the maximum rate at which Halon would dissolve in water to replen:'h that consumed is estimated to be approximately 20 pounds per hour. The value of 20 pounds per hour derived from consideration o' the rate of Halon dissolution is consistent with a G (-CF Br) of ~ J.5. Conse-3 l, 96<
l l 1 .\\ 7.- m, - -
- ' m
) ji i 1i. s. quently. 20 pounds per hour of Halon probably represents a likely maxi.. mum credible rate of dacomposition at the time of Halon injection and would decrease as the f;ssion product activity subsequently decays. De-composition of Halon, for an assumed G (-CF Br) of 0.5, is snown in 3 Figure R to illustrate the extent of decomposition for a credible and I i' i (pr'obably); conservative rite of Halcn radiofysis. These values, however, are subject to considerab:e uncertair.ty, and a more definitive evaluation cannot be made until (1) further qs.=ntitative information becomes avail-a',le on reactions of Halon with radiolutic decompositien products of syster, and (2) more definitive descriptions of the sequence of events and modes of system operation following a LOCA have been deveinped. i l 4. 2. [!adiolvsis of Gaseous Halon for in-dvertent iniection For the case of inadvertent injection of Palon into the reactor contain-1 : j ment atmosphere during normal operation, where continued oper4 tion is mandatory, only very minor decomposition of Halon is expected. II Simple calculations indicate that only 0.00023 grams per day cf Halon i. would be decomposed for each R/ hour source of gamma radiation to j which the Halon inventory is subjected. Preliminary information
- L indicates that the Halon atmosphere might be exposed to radiation fields of 100 R/ hour in limited regions of the containment vessel. Using the f
G values for the radiofysis of liquid Halon, and assuming the entire in-ventory of Halorr were exposed to a 100 R/hout source, a total of only { .023 grams per day of Halon would be decomposed. Even in~ creased by a large factor for uncertainty, the not decomposition of Halen would still i be quite smali for the postulated abnormal condition, i.e., inadverte9t injection during normal operation. i The small magnitude of gaseous Halon decomposition under the postulat-ed conditions is not unexpected. Extensive studies 4 of the radiation f chemistry nitrogen and oxygen have shown G values for nitrogen fixation (as various oxides) to be of the same order as the G value observed for i l pure Halon decomposition. Since the formation of nitrogen oxide (or nitric acid) in reactor containment atmospheres during normal operation I~ l. Tentative information provided by Atlantic Research Corporation. 1 974
i _ ~. -. - -. -. - - e _._ mwm wommrwoes cmew.r.w .-- - +- s i !I .~ I % '03SOdWO330 tWH i i t i O O O O O O f I 10 9 M N L h i .$a.a. .a._.. ~ .... k ,. 4 Lg. y. 7. J y p. .4, * $... 6 '6 .. oc f .l... 4.. J.] h -. g.'..-.., . 4.1 4 j.;.... 4: ,+4.,.. 4 L -64 . 4 4..4J41 - -- ~4.. 4 y. L.t, .e . 4. 9... L a...p4... ', + 4.g .q 4 p.+, +..; 9..-..tT**'t'..'~*.~.* . i. 4' "... a 6{,-...6 y.L .). ..r ,9 . A {.1 d
- * * ~ ~
.3.L.. 4 o. .. i J.... l !,.d... .6 .. L e.. ...s ..J.....1 i pa.. ...6 .l.;..l. .. g.1 s.... M6?,. L.-.... p+44 .. '.t**t ..7
- i. -*-
L -, 44 4 4.... s. 4 4.. 4 4 y &4 .44, ,,.e.4 L4 F..
- 9. l..... }..
- 4. a.L L....L.6.
A p. .4 4.J. 4}.. . 4 .1+i;4 4. .i. 9.. 4 L 4 8 L. -* -b'-. _.u+i--*'..*.*...-.-"&,"..- r. 4 L 6 o en f 6. t.*..e .p4 ..+.p ..:.+. p,..,.. ....9-.**M. ,a ; 64a - yy l! I!! 4 ,.4.- a.* a. f 4.; :~.La, L 4,: 6-- 6 j.,., e 4_ L,..h.. . y...} y.,' 'L.. 44 J L.,. qa ..s d 6.d.
- 44. a 4.
-+ 4L.. 3 .3 , +q i !l 4..-.,.L -... + ,a., 1 !>i .4 47. ' 4.4. 4 2 2 - -p p.J &. 4. 4.... h- -.. _. 4 -.; k. . 44 _e. ... 44 L..a.} :
- l g &.4,
4 o 4 L... .3e..... p.44 . g +... ,1 ."g .+ b.. ,j 4 4{.., 4 4. -. I l' . t.L .u .p. (..L..t.l 6-t,- p7.4 .,j 6. m
- 4. p.9 +.4
+. g
- 441, 3
.e.. . en .o }. 9., 4. . +. 4 L.. _ d. o.-. J a. p.,. o f A.L+ 4 A.o 4
- g 3 4
.4. -.f. . 4 4.4 . i 4.y .. "--'19 4 4 t U g . Li. a q -..... .t t-* L4. 4. *-.
- _-.-' 4. 7g.*4.-.I..
g F4 d . 4.4
- p... -
u.... - t. u .*.4 p. -. p ., l 7 s.. 4.,. p.. 6.. cg +,47..- + p .,y.,.+ ,g g l y.......... , 4 W.. ,,A.. ..}S.- 2-gg .,...a. g.......... ,4.. 7 ..,,4- . i.1,. 4... a p.4. 4 .g, &4, .. g..L.... - .}...v. .L ,9 E .. j p 6. t .t ~6.6 .r.... - +..o.. .6 ...; a ;. .. v 4.p. 49 .[. u 6o. .i....y... .g,..7_ .p.,... 4 F A
- b.,
.a ..L .p.-. ,.p.., 4. ,.....a.. g.,. i a.p I. a e.. L.... a 6. . ~. 4,..%. 5. 6. ... g.. a44,- .F.p6 }- w. 4 L, t.4. .. }. p.. .)4 L.. ..pt..h 7. . 7,+ ag-y.. - p p.4. %+ t .p. q ;m p9 . IM'.4 m......i-4.t,a 4.L' J4 ' t...*4.-' ' +4.. ' a.*.a "*; ' +i. M-*4i.8 4 -* ** b*-i.p 4* +* a i 4.
- aww...
1a 4.,.u ,e. 4
- e. a aa...
4a L..... 6 (.. }.. 4 ._4... ..a a L
- 7. j....p....'...
a4 pLu L ... *. -.. - +.. * ' _ '.. .-.-..-t.*s..+..+...**y.. .. f. r, -t..
- +
pli.b: . a 4x +, ..'...1 4.#.. t . 4.a. .& 4.
- .a,
_. g i p ., Lp? l.
- r'
. L4. aH
- 6"-*-"a..~,,...-.M.*-*--+..--..,'.~.1..*..4-*-+'
.t t a +- + 4 . 4.i. m........ -. a.. . ua.4 .4 4 ed. 6.,..- ..v._ -..p4 p 4. 4. a j,. 2. t.. 4 -
- ,, y s
.... ~~y 4,a , -a 4 l .-.4 '., 4 ' p, 6 +. ' }+ .p# - .+ M. ,4. 4.. 4 M wii i1. i i .I a s o 8 o o o a i i .n .n n i .o n .a n ~ 1 sqt *0350dWOMO NO"WH l 4 G.e 7 s ej 4 .m +-
~ ~~ _c_ I has not been a significant problem over long periods of ticae, neither i. wouki Halon decomposition be expected to be important, especially over the short penod of time of the postulated abnormal condition. If it is 1 assumed that the presence of oxygen, hydrogen and moisture could result in the proquetion of hydrobromic acid. with a G yield as high as 'a, (insteed of the G of 0.22 observad for elemental bromine), the total yield in 4 days of mandatory operation would result in only a small con-centration of HBr in the containment atmosphere (~ 0.02 milligrems per A-cubk: meter). This fevel is net expected to have significant consequences, althoug% the low yield and negligible consequences likely will require ex. I perimenta! confirmation. t. { 3. Excfws:e Reaction in accrtion to radiolytic effects on Halon, there is some possibility of an {; iodine reaction with Halon (perhaps radiation catalyzed) to form an organic iodids (trif!aoromethyl iodide), which might not be as easily re-moved by chair.oal filters as elemental iodine. The relative halide bond strary:ns of CF -Br and CF -1, however, suggest that high temperatures 3 3 (high activation energy) would be necessary for the exchange reaction to proceed at apprec~iable rates. Furthermore, measurements, 6 of the 5 equiiRbrusn constant confirm that the reaction does not favor formation f, of the or-anic iodide. At LOCA temperatures, the rate of the exchange v react >on would normally be quite low, perhaps requirirec, years for *:gni. ficant exchange. However, if radiation catalyzes the rate of the exchange reacten, the mass offeet of the large excess concentration of Halon could concehatly convert an appreciable fraction of the iodine into the organic i form. Th:s effect warrants further investigation, principalty to allay any concern and to confirm the anticipated absence of significant organic iodide fvmation. C. Recome. wrier,s it is recommended that axperimental investigation of factors affecting the radiolytic cocomposition of Halon be undartaken to provide the basic informa. tion necessary to reliably and cuantitatively assess the magnitude of the M-tl.. 1, 99< n. e 9 D. 3 9:
i ( possible problem. The major aisas requiring experimental investigation are list-ed below in a suggest6d order of pric,rity, established on the basis of their signi- { ficance to the proposed use of Halon for combustible gas control and, the {' magnitude of the potential problem. (1) Radiolysis of acueous solutions of Halon to establish the G values and the mechanism for decomposition and to determine the products of radiolytic 3 decomposition, These tests should be performed under steady state 8 l irradiation for sufficiently long periods of time (~ 10 Rad) to determine f 'l macroscopic yields of products and the effect of accumulated reaction {J products on Halon decomposition rates. Experimental determination of the rate at which Halon in solution may be replenished (i.e., rates of dis-( g solu.lon) is also needed as part of this effort. d 'j (2) fodine exchance reactions with Halon should be investigated expe'ri-5 j mentally to determine if significant conversion of CF Br to CF ' 35 3 3 possible in the presence of radiation as a potential catalyst. These tests should $ made in both the gaseous and solution phases and should ~ include tests with the bromide ion (and perhaps fluoride lon) present in 4 { solution to simulate ac:umulated oecomposition products of aqueous I. Halon radiotysis, in addition, tests of charcoal absorption of the organic iodide (CF 1) form are also needed, particularly if significant formation of 3 ), CF 1 is found m the exchange tests. 3 (3) Ges-ohase irrediation of Halon, alt, hydrogen and water vapor mixtures !, { will ultimately be needed to confirm the expected low rate of decom-position and to demonstrate the absence of any chain-reactions that might {~ lead to unexpected Iarge decomposition rates. i (4) Depending upon results of the tests recommended above, some experi-mental investigatien of the metallurgical effects of Halon decomposition products may eventually be necessary. \\ I l 100< s
l lil. LOSS OF CocteN T ACCIDENT SEQUENCE The design basis accident 7 for a loss of coolant analysis is a break of the 5" surge line, which is the targest line in the reactor coolant system (RCS). While the tsactor cool-I ant system is blowing down, the containment pressure increases, rerhing & peak of j [] 68.5 psig at the end of blowdown spproximately 300 see after the break occurs. } The high pressure injection system comes on when the RCS pressure drops to 1600 { psig and the containment pressure reacnes 5 psig. This system, which is normally the makeup and purification system, pumps water into the RCS at 150 gpm (two of three pumps) from the emergency storage tanks (two tanks,74,000 gallons each). When the ~'. reactor coolant system pressure drops to 200 psig, low pressuru injection (normally the decay heat removal system) is initiated at 500 gpm (ene of two purra) from the emergency storage tanks. After the storage tanks are emptied, the recirculation sys-tem (LPI system) takes spilled coolant and injection water from the containment l sump and retums it to the reactor coolant system after cooling the water through the 1 deczy heat rarnoval coolers. Heat is also removed from the secondary side of the If , steam generator during a LOCA by the steam-turbine driven emergency decay heat i pumps. These systems are shown schematically in Figure 2. i e ) e 4 I !~ l i i 1 1 101< 1 j
-__~ ~. ECCS SYSTDIS OVERBOARD IL B EMERGENCY WATER STORAGE t =~ E.5Q O~X ~.~. ..v.' ~
- I f '-
NORMAL I o o' i L, 5 Y MAKE-UP AND* REACTOR 4 O PURIFICATION PLAitT I HPI SEAWATER l .? I ~ \\. 1 'r NORMAL / POST LOCA T $g STEAM --C><} COMPONEtli-3h M I FEED NORMAL COOLING e --{><} 4),.; Eh- -r A I _\\ WATER .b N Et1ERGENCY -(" DECAY llEAT ) 110Rt1AL REl10 VAL DECAY HEAT + ~ ~Q ,H NORMAL t REMOVAL l U LPI + OVER3OARD + i POST LOCA Ef4ERGENCY MATER ST0 RAGE ll.;. (. ~.:'.. :.,..'. : * ;:.f..:.' :j ':. REACT 0R PLAflT ~ SEAtlATER f LONG TERM CORE COOLING EMERGENCY EMERGENCY WATER STORAGE r-- f Fig. 2. Schematic of Normal and Emergency Cooling Systems 5 4 +
Il t L. l t_ IV. RADIOLYSIS OF HALON. AIR MIXTURES One of the possible abnormal conditions to be investigated is the inadvertent opera- ) L-tion of the Halon injection system during normal operation of the reactor. For this conditfort, it was assumed that the Halon is injected into the containment atmosphere n;, at a time when other circumstances (e.g., a hurricane) make it mandatory to continue operation of tha reactor. In the absence of experimental data on the radiolysis of l was used. Yields and G gaseous Halon, yield data from Irradiatlon in liquid form CF (G = 0.62), C F2 6 (G = 0.25), C8'2 2 (G = 0.98), and F values from Ref.1 are: 4 { Br2 (G = 0.22). With oxygen, hydrogen, and water vapor present, the products of the ! " ~' radiolysis of gaseous Halon could possibly differ from those reported in Ref.1, with j some HF and H8r produced and probably lower yields of the higher-molecular-weight j <.h compounds. Despite the fact that bromides are notorious for creating chain reactions. there is no obvious reason or justification for postulating the existence of such reac-( tions. Nevertheless, it is likely that experimental confirmation of the absence of chain
- ij reactions with high G (.CF 8r) values will eventually be necessary.
3 J l in the absence of data on specific radiation sources within the containment vessel, it has been assumed that all sources are located near the reactor vessel and that radiation passing through the containment atmosphere will traverse a distance of 12 feet
- before entering the outer radiation f,hield. It is also assumed that inadver*.ent operation of the i
system injects the entire Halon inventory m storage, although redundant storage tanks i may reduce the likellhad of injecting all the Halon. Taking, as an example, a tenta-tive Halon inventory of 3.840 pounds, the density of Halon gas in the con *ainment g atmosphere would be approximately 0.0015 grams per cc after injection. Using a 7 2 mass energy absorption coefficient of 0.03 cm per gram over a broad energy ~ spectrum, approximately 1.66% of the radiation would be absorbed by the Halon gas, j For a total G ( CF Br) of 2.1 in the aseous phase, each R/ hour at the source would i
- o. 0 00 L 1 fpa decompose approximately.Q.OOPf3' grams of Halon per day. Tentative information from Atlantic Research Corporation indicates that local sources of as much as 100 R/ hour may exist. If the entire inventory of Halon in the containment atmosphere were exposed to 100 R/ hour, the not decomposition would amount to approxiraately I
0223 grams per day, ~ Scaled from information and drawings Oven in the PSAR.8 j' 1. i, 103<
~ = ~ i e l -.t I L . c j j For the abnormal condition postulated (i.e., inadvertent operation), the total decom. t L position of gaseous Halon l'n the containment atmo;phere would not be sufficient to constitute a maior cause for concern, even if higher radiation sources existed and con. [, tinued operation was required for a period of 4 to 10 days. For example, continued operation for 10 days with a 100 R/ hour source might decompose approximately 0.23 grams of Halon. i The potential for halide stress corrosion appears to be the only adverse effect under U the postulated abnormal condition. Although no information is available on the yield of hydrebromic acid, if a G (H8r) of 1.0 is assumed (probably conservative), exposure ' {b of the Halon in a 100 R/ hour source for a 4 day continued operating period would l, result in the production of approximately 024 grams of HSr. In the total contain-i j ment free air volume, this would amount to approximately 0.02 milligrams per cubic meter which is well less than the recommeridod9 limit for continued human occupation 3 (10 mg/m ). No information has been found on the effectiveness of the bromide ion in halide stress corrosion. However, if it is conservatively assumed that bromide ion is as effective as the more reactive' chloride ion, the bromide ion concentration in the d containment atmosphere it is still too low to be of serious concern. In the absence of l specific information on radiation sources and actual yields from rMiolysis of gaseous 3 e I L.3 mixtures of Halon-oxygen-hydrogen water vapor, the evaluation given above can only i i j b? considered tentative Although no problem appears to exist, experimental con. 1 firrr.stion of the radiolysis yields is needed (i.e., to demonstrate the absence of any I chain reactions), and further evaluation of the actual radiation sources which may ] l l exist will be necessary.' 1 J l t t ~ j
- \\
j. 3 j i t. 10
- An earlier evaluation of the radiation source from oxygen activation in the main steam line of the CNSG reactor incicates a much lower radiation source than that used in the evaluation above.
l r l 104< 3 n ..c ---m- .--,.,e,.. e---.-- .-~.e
~ + z t V. RADIOLYSIS OF HALON IN SOLUTION (' The principal mechanism for the rad:olytic decomposition of Halon under post.LOCA b conditions is postulattd to be interaction with products of water radiolysis. Because e Halon is only slightly soluble in water, the fraction of the decay energy absorbed direct. L ly by a Halon mole:A h too small to constitute a significant mechanism for decom-1 position. Essentially all of the radiation would be captured by the water, and the j resulting water decomposition products would then react r'.nically with the Halon in solution. The solubility of Halon under conditions expected to exist in the con-tainment following a LOCA is estimated to be approximately 0.08 grams par liter or J approximately 0.00054 moles per liter. This concentration is still relatively high com-4 to 10-10 moles per pared to the approximate concentration levels (of the order of 10 1 6! liter) of free radicals in solution from water radiofysis. 1 A. Halon Sctubility 3 Under equilibrium conditions, Halon has been reported to be soluble in water I 0 { to the extent of 0.0003 grams per cc at 77 F and 1 atmosphere pressure. The quantity of Halon dissolved in water under post LOCA concitions would l be appreciably less because of the higher tamperature and lower pressure that would exist. In addition, following tho injection of Halon into the containment j atmosphere, some time would be required for the Haion to dissolve and the b concentrafien to approach equilibrium. For post LOCA conditions, the
- p approximata concentration of Halon in solution at equilibrium is estimated to j
L. be approximately 0.08 grams per liter, assuming a Halon partial pressure of j ~ 4 psia after injection. Assuming this value is appropriate, the equilibrium amount of Halon dissolved would be about 9.3 pounds of Halon in the wetwell water and about 3 pounds of Halon in the primary system and sump volume. At the time following a LCCA when Hafen is injected, the approach to an { equilibrium solution in the wetwell water will be strongly hiedered by the absence of any mixing or recirculation mechanisms and the relatively small
- Following the LOCA and iniection of water from the emergency storsgo tanks, the wetwell I
is estimated to contain approximately 18,630 cu. ft. ano'the drywell, including primary system and sump volume, to contain approximately 6,400 cu. ft. of water. The total volume of water spilled frem the broken surge line and collecting in the sump is estimated to be about 1,890 cu. ft. (Derived from information in Ref. 8). I. 105<
-r j surface area of exposed water. Consequently, any Halon consumed by radio-lysis in this water would not be. easily replenished. I ;I For tne primary coolant and sump water, operation of the low pressure inject. I ! ion system would recirculate approximately 500 gpm of water from the sump ) mto the primary system, spilling out of the line assumed to have broken in the daugn basis accident. Assuming that this recirculation flow (500 gpm or 30,000 gph) would attain a saturated Halon concentration as it splashes thrnugh the containment atmosphere from the broken pipe, the mnimum instantan-l [. sous rate of Halon dissolution would be approximately 20 pounds per hour. Consequently, even if the Halon in solution were totally decomposed during \\g traverse from the sump through the core and out the broken pipe, the overall l Halon decomposition rate would be limited by the rate of dissolution in water. , "j Thus, an upper limit on the amount of Halon radiolytically decomposed of j approximately 20 pounds per hour would' be determined by the rate of dis-solution. At some time fator in the LOCA accident sequence, the recircula-L tion mode o.f operation can be essentially terminated and cooling provided by the emerguncy decay heat removal system or by the ordinary decay heat remov. ~ al system, either of which can avoid spraying coolant into the atmosphere from the broken pipe. Under these conditions, the net decomposition of Halon could be significantly reduced and would then becomo lim lted by the diffusion-controlled rate of Halon dissolution on the exposed surface of the sump water. B. Radiolysis of Water Recently, there have been a number of sutveys of the radiolytic decomposition l, of water and of the highly reactive intermediate free radical products formed.11,12.13,14 The radiolytic decomposition of water is generally de-scsibed in terms of the ionization of water and the formation of transient ~ reactive intermediate products or free radicals. In absorption, radiative energy i is lost to the water, creating a local region with a high concentration of ions and free radicals within the track length of the ionizing radiation. Within a very snort period of time (of the order of 1010 seconds), local interaction between 3 these products has been essentially completed and diffusion away from the track (sometimes referred to as a spur) quickly dilutes the concentration of free radicals and greatly decreases the probability of further interactions. i 1 i 106< l
o ,W n e 1 6 Thereafter, reactions occur predominantly with other materials present in the system. Products formed within the immediate track length and diffusing into the medium are the free radicals H and OH as well as the molecular products-j H and H 0. Two apparent forms of the H free radical have been identified, l 2 22 l differing significantly in reactivity. One of these--the most reactive-is gener. f ally identified as the " hydrated electron", or er",q while the less reactive form is { (H). Observed yields of the various products are relatively insensitive to pH, and in neutral water are approximately as follows: iI
- t e*,q (g of 2.05 2.7), H (g = 0.S.O.6), OH (g = 2.6 2.7),
i H02 2 (g = 2.7), and H2 (g = 0.45). l .) I interactions of these various materials in pure water result in overall net s observable yields of hydrogen, oxygen, and H 0. For convenience, different 22 nomenclature is normally used to designate the two different types of yields: l the immediate or primary yields of free radicals and molecular products are designated by the letter g, and the overall net macroscopic yields observed are designated by the letter G. The primary or microscopic yields, g. are not appreciably affec*.ed by the addition of most other materials in low concentra-tion. However, macroscopic or not observaW yields, G, can be significsntly { affected by reaction with various chemicals present in solution, even in low concentrations. Such is the case with low concentrations of Halon present in I solution; this accounts for the postulated mechanism for radiolytic decom. ( position of Halon. In the absence of any materials in solution, +.he principal reactions in the radio-lysis of pura water are the following: j HO a;q,OH,H.H0,H2 (1) 2 22 t + HO -H + OH" k = 16 fE sec.1 I (2) e;q 2 + H0 OH + OH-k = 1.2 x 1010 pg1,c.1 (3) e;q 3 22 1. t-1, 107<
s __... I L (4) H+HO - if + OH k = 9 x 10 u-I WI 7 2g W sec'I k=2xIEO I -HO2 (5) H+O2 H O + HO k = 4.5 x 10 W sec 1 7 I ~ (6), OH '+ H 022 2 2 l 7 I I -HO+H k =4.5 x 10 W sec (7) OH + H2 2 With no other materials with wench to react. the very diikne reactiortproducts eventually interact with each crimr, for example, 1 -: J WWI I z + OH CH-k=3x1SO (8) e i <j W ac I I (9) H+H -H k=2xidO 2 1 (10) OH + OH' -H 0 k=0.5xly0g1 g1 22 i W sec'I I (11) H + OH -KO k=2x160 2 ] (12) 2HO2 k=3x19 W mc"I I H022+O2 HO+O k=4x19W WI I (13) OH + HO2 2 2 For the radiolysis of pure we:ar, the macrescopk or rut feid of hydrogen, G (H ), is quite small,large y F=-= of the recomtse.at' c reactions occurring 5 s 2 in solution (see equation 7 ateve), despite the fact t521rie microscopic yield, g (H ), is 0.45. However, if scme material is preserri est reacts efficiently I-2 with, or scavenges, the OH raica!, recombination is h= and the net rr. acro-reopic yield of molecular Mw.G (H ), approaches ee microscopic value, 2 g (H ) of 0.45. 2 In Regulatory Guide 1.7, the USAEC states that a yhd nka G (H ) of 0.5 2 be assumed when assessirg ecs:.LOCA radiolytk h:::::r;en accumulation. t l,' 108< l l
For the Maritime reactor, calculations of the time dependent hydrogen concen-tration,* using the assumptions specified in Regulatory Guide 1.7, result in the hydrogen conantration history,hown in Figures 3A and 38, for radiolysis only. To prevent hydrogen cornbustion, Halon injection is required when the hydrogen concentration has reached approxirnately 3.5% (allowing a margin of 0.5% below the specified cembustible limit of 4% hydrogen). As indicated in Figure '4B, Halon injection is required approximately 6.6 hours 1 ~' after the occurrence of the postulated LOCA. Injection of Halon results in a'i immediate reduction of the hydrogen concentration. At the present time, the i Halon is expected to be mixed in the containment atmosphere in a few minutes, j although a considerably longer time is available to insure uniform mixing of the i atmosphere. The case shown in Figure 3B assumes, for illustrative purposes, the f injection of 7:150 SCF of Halon.** With this quantity injected, the Halon con-centratton at the time of injection (6.6 hours) would tl,e approximately 37 volume per:ent. For the case illustrated, dilution by the Halon would allow up f to about 10 hours to accomplish uniform mixing in the atmosphere. Thus, early in t te accident sequence, there would be more than adaquate Halon present f a completely suppress ignition and to minimize potential problems of mixir g. The actual quantity of Halon and method of system operation (i.e., whether injected in a single step or as a series of injections at various time l t, intervals) remains to be determined later in the detailed design phase. To some extent, the procedure for Halon injection ultimately adopted can have an of feet on the extent of potentiol'radfolytic decomposition, because of the quantity i a that may be pr:sent estly in the accident sequence when radiation levels are j higher. L @p j, , Calculations performed with the SNLH2 code developed by NUS Corporation, Detailed { t, printouts of these calculatierts are attached as Appendix A.
- This quantity of Haton a559tned is the amount needed to provide 20 volume percent Halon at a time later in the accrder.t sequence when the hydroepn has accumulated to 40 volume percent.
} 103< 1
e e t s,_. L- - i-s. L._.. L--. l.. - L- . I I I 25.0 1. p.. 2 3i ......s_..,... 4p.#..:. - g_ t ...+,.. t.... p., - l t ......9 4 . 4,. 4 ....,j a ; '.. 3 .I 1 7 .. :.. T --. r- ' '-]-' M[ 8 * } -t 20.0
- e--.s 7
- p. _M_.,_:_e...
1 1-44 M _,: 4 .L. .l... q g.y_ n :_g.:.
- q...
_r.f ... v.,. _ L -t-g.7 _ .7.. 3,. a t.... .r-- ,I- . c
- - '- '
- r t -4 7.'
1 r r -- - 15.0 f, ,,T t g4 . g g,p ..A. _.4 g4 . t- - t _..7._ jp .--{- i y A.
- 3.,
p m. ,, y p . 3 ,.7 n ,op-t ~ Y X .. r.'.' L 4 i,- - ^ - -....-.-~l~~' .'.-..-.;-*2 "4'.i- - ' - +l-4 L.- ". '... -. ~. _ ' '. '. ". _ ' ' '. ' I- +r L{ %. 4 , 7 ..a.. C a 10 0 A _..i-+_.. .t: r :-.- - - ~ f = I' ' ~~ t - - - 7-O ~* . '. '. '. ~ ~ '..~~'..7.', .'..M.-.t_.'., ~ l]-I. ' L~' ,,. g b y.'*
- '.'s..,'..p.,'
I'I T "'~ s... .4. .4 .,.. g9. .,.pg 3 .,1 .y;.} - g :., 4 !..).&g,*. a 4 ~. 8, r... .g. 9 a g .t... ..,. leg . t.t l.- t
- t.:
t
- q.
5.0 +---- 1' : .l .i;l I ..,. 4 4.- }. 1 ..,,i g g.... L ....., g g.. .i,i,.. .,, 3... g :.. .g ...... }.. g g.. }.. pj........ ..i...s.t g..a;.,.. 9.:.;.g. .'..a.
- ... 6...
,g... 4 . g.;. s. su. ...a. -,.g. IL(.
- -I.,.. --! l I 4 g....
e..... I,.{ $ ,g.. i .a.. 1 i r, j.1 - l. .) g j. ~ . 1.
- rtt::+
tt;t't:ir 1 t-it l +-r- -T' 3 0 0 25 50 75 100 .125 150 175 200 f t TIME. HOURS g Fig.3A. Accumulated Hydrogen Concentration in Containment Atmosphere (Radiolysis only. Reg. Guide 1.7 (Versica 2) Guidelines) { b Oa. 4
+ .. ~.. n l y n I e a s 5.0 l .l .p.. --.-..-.m-g
- 1.
7t. . - * - + .t. - : -...- . -. + .g 1-p t .pL y-., M, g .. -.., - + + . 44 pg. 1. -.7 ; *J* -s L.r ': ^ 4 -t , _ p., .y pl-. 4 4.0 .,.3 t
- ,t.,
.j... /gl }
- h.,,a, ? g. a.,
f. I!,,(p 7. - j 1.1 fit .1 . y 6 . t.. }, s.. a [. 4 4,... u... 1 _l t-*****' 4 { w., %. 7 3 Ip 2....26 ...~ .1 g. .-{.. ---7 d, g ]'O '.,H -{T ,s tJ.1 ) --.twp . f4 II..g I, 3./'p I -.. .. a -4 y n'u.0fE11 KCI DIf ' #d i .0 r, .~ {j .,t.- y4 G.$ 4_ns, .l . a.-
- 4..
...g .i, f ,7.,., x
- r. p..,,-, -. -}. n.
.i n .-a.7 ._+- --. ;..t..,...q i 1 s 2.0 p [. ......g, 3.1 4 - l.6. -(
- e. 9.
m... , f ghM N F-r-* ,... 7 % s f 4 f _.%. _t. -.,- l .. a g4 ,r r. ly A .....g., ..... p. t.. j.{Q- ~7 10 9..g j j.4 a '.L i. --+- ..s....., ...,7...
- 1. 1,
T... 1- ... ~.... .....),... s., 1 -g . _f_, .... l.g ... :.1.i. .1. .r. u ..qd 7. .= l . l .. r.,,7 0 0 5 10 15 20 i TIME HOURS I Fig. 38. Accumulated Hydtc9en Coricentration in Containment Atmosphere (Radiotysis only. Ret Guide 1.7 (Version 2) Guidelines) 1 1 I r f E
- -.~. 4 L C. Halon Solution Radiofysis Unfortunately, there.re no published data on reaction products or yields from steady state radiolysis of Ha'on in aqueous solution. Bullock and Cocper 2 in l pulse radiolysis studies of trifluoromethyl radical reactions, have used the reaction between the soivat*<8 electron and bromotrifluoromethane in solutio's ,l as a source of trifluoromethyl radicals. The short-term yield is apparently quantitative and results in the fctmation of the trifluoromethyl radical with a g value of about 2.65 to 2.7. The subsequent fate of the trifluoromethyl radi-i ~ cal in the absence of othe chemical agents is unknown. However, the initial l reaction with tN. solvated electron is perhaps the principal mechanism for J radiolytic decomposition or Halon in solution. Bullock and Cooper, uMng spectroscopic techniques, did not observe an absorption band attributable to l the carboxyl radical although absorption due to the HO radical was observed. s 2 l Consequently, carboxyl formation and hydrolysis, analogous to decomposition I by radiolysis of some chlorine compounds, does not appear to be significant in Halon radiolysis. Uapublished data of Cooper 15 indicate that bromide ions are present and that the r.tacroscopic G value derived from the OH radical attack en bromide ion is approximately 2.5. These data support the expected g ( CF Br) of 2.65 to 2.7, 3 with the CF3 radical and bromide ion as the principal immeciate products of L aqueous Halon radiolysis. Bromide ions have previousiy been reported to be quite effectiveI3' 14 in i reaction with the OH radical formed in water radiolysis. As bror. tide ions l accumulate in :.olution, they will effectively scawge all or nearly all of the OH i radical with two major corsequences: '1) recombination of mulecular hydrogen by reaction with OH will be pt vented, and (2) decompositica reactions of CF and OH will be affected. Under these conditions, radiofysis of 3 Halon solutions would tend to result in higher molecular hydrogen yields, with i a macroscopic G value approaching that of the microscopic value, g = OAS. Since this limiting value for the macroscopic yield of r'io'ecular hydrogen has { already bean used in assesng the concentration history of sceumv8sted F 112<
t 4 ) hydrogen, the net effective yield of hydrgen gas will not be increased. I Bromide scavenging of the OM radical wi!l, however, compete with the CF3 radical reaction for OH. I The presence of bromide ions is not expected to create a chain reaction. I Possible chain rea'etions, as indicated in the following, de not apoear likely. ~ t (14) CF Br + Br CF3 + Br2 3 i-(15) CF3 + CF Br CF2 6 + Br 3 v. The first of these reactions is endothermic and, in addition, Cooper 15 has PJ' observed only a small yield of C F (G of less than 0.5). Consequently,if the 26 potential chain reactions occur to a limited extent, they do not proceed very 8(. far before being terrninated by other reactions. Little or no not yield of CF H has been observed 15 indicating that the follow-3 ing potential reactions do not occur to any significant extent. ^ 'T CF Br + H CF H + HBr 3 2 3 i j (17) CF3+H2
- CF H + H 3
(18) CF3+H022 CF H + HO2 3 CF H + O (19) CF3+HO2 3 2 ) i Fluoride ions have been reported 15 to be present, but no quantitative data is currently available. Fluoride ions could originate from a transient very rapidly hydrolyred carboxyl radical of through formation of CF 0H. No detectable 3 amount of CF 0H has been observed and, if formed, probably hydrolyzes 3 rapidly to cart ;nate. 113< l l
lL L t Irradiation m16,17 of the chlorino analog of Halon resulted in decomposi-tion of CF CI by dissociative electron capture, with a yield corresponding to 3 complete reaction with the hydrated electron, i.e., CF CI + e*,p CF3 + Cl* k = 4.4 x 10 M-1 sec' I 9 (20) 3 s 1 in addition, the CF radical reacted with the OH radical, 3 { HO [CF0H 2 CO + 3HF l (21) CF3 + OH 3 2 i f For the CF Br system, however, the bromide ion produced by dissociative 3 electron caphne, competes with CF for the OH radical. 3 i iJ (22) OH + Br-OH* + Br k = 1.2 x 10 M*I sec"I 9 L. Consequentty, the decomoosition of CF3 by OH capture and subsequent hydrolys3 may wsll be censiderably less with Halon then with CF Cl. 3 On the basis of the information currently available, it is not possible to state with assurwce the mechanisms yields, or products of the radiolysis of Haton in aqueous solution. A maximum G (-CF Br) of 2.65 to 2.7 would result 3 from the assurnptions of complete reaction of Halon with all hydrated elec-trons formed and complete decompcsition of CF. The subsequent fate of 3 the trifhroromethyl radical, howaver, is quite uncertain and the recombina. g tion reaction (23) CF3 + Br + CF 8r 3 could apcreciably reduce the net macroscopic G ( CF Br), especially later in 3 time as the Sa* ion accumulates in solution. The small yield of C F obsented 26 by Cooper, ard the (perhaps) small yields of fluoride ions may be indicative of a low over22 r-.acroscopic G value. 114<
1 1 L. I i For purposes of the present evaluation, the maximum G (-CF Br) of 2.7 could 3 a result in a maximum decomposition rate of 90 pounds of Halon per hour at the time of injection (6.6 hours following the LOCA). This rate would decrease l with time and at 35 hours after the LOCA would be approximately 47 pounds i of Halon per hour. If s!gnificant recombination of the trifluoromethyl radical i I exists. the rates for an assumed G ( CF Br) of 0.5 would be approximately 17 3 pounds per hour
- at 6.6 hours and 8.7 pounds per hour at 35 hours. These data are shown graphically in Figure 4 a
L. I } Despite the uncertainty in radiolytic decomposition rates for Halon in solution, I-the effective rates will be affected (and perhaps controlled) by the rate at which Halon in solution is replenished from Halon in the atmosphere. As indicated above, the maximum dissolution rate could be as much as 20 pounds per hour a (instantaneous value), which is only slightly larger than the solution decom-position rate resulting from a G value of 0.5. Consequently, the lower curve shown in Figure 4 is probably more nearly applicable. although still conserva-l tive for the radiolytic decomposition of Halon under post LOCA conditions. Clearly, additional experimental work is necessary to adequately describe the radiolytic decomposition of Halon. In addition, a more definitia description of the sequence of events, including proposed mechanisms for cooling following a postulated LOCA, is needed. Solubility data over the temperature and pressure range expected, as well as data on rates of solution, are also needed. w e 7 These values are for total decomposition in the entire inventory of water. Distribution of fission products, particularly released iodine, between wetwell and drywell could further reduce the amount of Halon decomposed depending upon the mechanism eventually found to be controlling. 115n
IJ t 18-i L., g 2 ..a o ~ i; a. I 2 a M z 10 O u ~ 2 i o G = 0.5 t .s 4z I u. ,J o ws ,1 j.) eil ,....I ,,I' i . i i i, e i ,. i,,,, I' 10 101 '04 104 kj TIME,(hrt) 1 6 '~ id ; k.s 4 ~ Il U ~ Gs2,7 a. b ..e m 2 8 10 'm 9 4 e r-I go ~ W Hd Z t t ? ? I t !#! t t i. .I ?,I i e i, e , i,,,. 40 102 103 10' TlWE, (hrt) Fig. 4. Time Dependent %ts of Psion Consumption i 116-
a
===r-l ,41 VI. CHANGE REAC,rlONS The existence of the trifluoromethyl radical in solution and the possibility for radia- ] tion catalyzed exchange reactions between Halon and iodine in the atmosphere should also be taken into consideration. Theexchange reaction r-J CF Br + 1 CF 1 + IBr O H = 9.55 Kcal/ mole 3 2 3 I has been studied 5 0 at temperatures around 700 K (8000 F). The equilibrium constant ( l 1 (CF 1 (IBr 3 K= = %.09 {CF Br{l) 3 2 t g i .j does not favor the formation of trifluoromethyl lodide, although the large excess in r ( Halon concentration could, by the mass-action effect, push the reaction in the direction f,i . of converting a significant fraction of the iodine to organic form. Tentative calculations [ indicate that, if equilibrium could be attained, as much as 90% of the lodine could be I converted to the organic form, which may be non-filterable. However, at the tempera- ~ { "" tures expected in the conta nmerit, the rate of this reaction is expected to be quite low, i-unless radiation fuoctions as an effective catalyst. In addition, the accumulation of Br* hd and Br2 n solution from Halon decomposition could also be effective in preventing i [ organic lodide formation. Whether extensive conversion would actually be experienced m. i! j under post LOCA conditions remains to be determined. k r i i,1 i . ~ ) i a t. s I 1. l' i' 117c i.
E 7fL h 7 REFERENCES i r o 1.
- l. McAlpine and H. Sutcliffe, A Comparison of the Radiotysts of Uquid Bromotri-
') fluoromethane with the Radiolysis of Uquid Trifluorpiodomethane, J. Phys. Chem., 76(15): 2070-2071 (1972). 2. G. Bullock and R. Cooper, Reactions of Aqueous Trifluoromethyl Radicals. Trans. L Farraday Soc.,66(Pt. 8): 2055 2064 (1970). 3.- DuPont Halon 1301 Fire Extinguishant, DuPont Bulletin No. B-29C. l. 4. P. Harteck and S. Dondes, Radiation CheInistry of Gases, Final Report, ansselaer Polytechnic Institute, RPI 32119, May 1969. i)~ 5. A. Lord, C. A. Goy, and H. O. Pritchard, The Heats of Formation of Triftuoromethyl l Chloride and Bromide, J. Phys. Chem.,71(8): 2705 2707 (July 1967). .I 6. C. A. Goy, A. Lord, and H. O. Pritchard, Kinetics and Therrnodynamics of the Rese. tion between todine and Fluoroform and the Heat of Formation of Trifluoromethyl l lodine, J. Phys. Chem., 71(4): 1086-1089 (March 1967), 7. Radiological Health Handbook, U. S. Department of Health, Education and Welfare, i 1970 Edition. 8. Preliminary Safety Analysis Peport, Nuclear Merchant Shrp Program, and Transcript of ACRS Subcommittee Meeting of October 3,1974. ( 9. N l. Sax, Dangerous Procarties of Industrial Materlats, Third Editinn, Reinhold Book j Corp.,1968, 10. F. N. Anderson, CNSG -Nuclear Ship Oxygen Activation Calculational Method with I Experimental Comparison, in National Tooical Meeting on New Dewlopments in I Reactor Physics and Shielding. Sept. 12 15 1972. USAEC Report CONF 720901 (Book 1), U.S. Atomic Energy Commission. I 11. S. E. Turner, Radiolytic Decomposition of Water in Water. Moderated Reactors Under t 1 Accident Conditions, Reactor and Fuel Processinq Technofocy. 12(1): 66 (Winter j 1968 1969). t i 12. E. J. Hart and M. Anbar, The Hydrated Electron, Wiley.interscience, a division of i John Wiley & Sons, New York, N. Y.,1970. 13. J. K. Thomas. Elementary Processes and Reactions in the Radiolysis of Water, in Advances in Radiation Chamktry, Vol.1, pp.103198, edited by M. Burton and J. Magee, Wiley interscience, a civision of John Wiley & Sons, New York, N. Y.,1969. 14. M. Anbar, Water and Acueous Solutions, in Fundamental Proe*ces in Radiation g Chemistrv, Chap.10, pp. 651685, edited by P. Austoos, interscience Puoissners, a division of John Wiley & Sons. New York, N. Y.,1968. l. F 118< l.
i, ) , _\\ i L ~ REFERENCES (Continued) IS. Private communications, R. Cooper to S. E. Turnw. ~ 16. J. Ulle. D. Behar, R. J. Sujdak, and R. H. Schulw. On the Lifetime of Trifluoromethyl Radical in Aqueous $olution, J. Phys. Chm, 76(18): 2517 2519(1972). L 17. T.1. 8:lkas, J. H. Fendler, and R. H. Schuler. The Radiation Chemistry of Aquecm1 Solutions of CFCl, CF Cl, and CF Cl, J. Phys.' Chem., 75(4): 455-467 (1971). 3 2 2 3 i j L i = 5,.' 6 e i_ E Ii4 i [It. i e 0 ) - 119< t I
.- - - - g - - -._ ,.i s APPENDIX A POST-LOCA HYDROCEN CONCENTRATION (SNE2 Code) j SN E2 Code to Calculate Hydrogen and Orygen Accumulation in Reactor Systems Following a LOCA in CNSG o
- t i[
INFITT DATA AND ASSLHPTIONS I e-h Fraction of 1 Released to H 0: 0.50 2 2 Fraction of Xe and Kr Decay Energy Absorbed in H 0: 0.00 2 J Fraction of Solid EP Released to H 0: 0.01 2 l f.. P Fraction of Gamma Radiation frem Fuel Rods Absorbed in H 0: 0.10 2 C (H ): 0.500 2 ('. Reactor Power (MW): 319.0 Veight Zr Cladding (1bs): 0.0 Containment Air Volume: 40742 cu. ft. at 5 PSIA and 120'F (Initially containing 20.90 )2) i i ANS Standard Decay Energy Multiplied by: 1.200 Percent Metal - H O Reaction: 0.00 2 6 k. k. 120< l
7._. p. p. OUTPUT DATA SNEH2 SOUTitERN NUCLEAR ENG. HERCilAITT SillP - ARC PROJECT 1-27-75 INITIAL IIYDROGEN .00 SCF I IIYDMK;EN FIOM HLTA1. - 0*00 SCF li 0 NEACTION g TillE PADIOLYTIC llYDROGEN OXYGEN SECS RATE, ScrH SCF VOI. % VOI. t 0.1000F 01 0.4614E 01 0.2307E 01 0.19641-01 0.2091E 02 0.1584E 01 0.4413E 01 0.2351E 01 0.2001E-01 0.2091E 02 0.2311E 01 0.4224E 01 0.2417E 01 0.2058E-01 0.2091E 02 0.39111E 01 0.4045E 01 0.2510E 01 0.2144E-01 0.2091E 02 0.6309E 01 0.3876E 01 0.2672E 01 0.2274E-01 0.2091E 02 0.1000E O2 0.3716E 01 0.2905E 01 0.2473 E-01 0.2091E 02 0.1584E 02 U.3527E 01 0.3257Z 01
- 0. 2772 E-01 0.2091E 02 0.2511E 02 0.3363E 01 0.3789E 01 0.3224E-01 0.2092E 02 0.39BIE 02 0.3209E 01 0.4592E 01
- 0. 3908E-01 0.2092E 02 0.6309E 02 0.3064E 01 0.5808E 01 0.4942E-01 0.2092E 02 0.1000E 03 0.2926E 01 0.7648E 01 0.6505E-01 0.2092E 02 0.1584K 03 0.2760E 01 0.1042E 02 0.8863E-01 0.2093E 02 0.2511E 03 0.2519E 01 0.1448E 02 0.1231E 00 0.2093E 02 0.3981E 03 0.2302E 01 0.2037E 02 0.1730E 00 0.2094E 02 0.6309E 03 0.2105E 01 0.2890E 02 0.2451E 00 0.2096E 02 0.1000E 04 0.1920s 01 0.4124E 02 0.3493E 00 0.2097E 02 0.158411 04 0.1740E 01 0.5903E 02 0.4989E 00 0.2100E 02 j
O.2511E 04 0.1561E 01 0.8447E 02 0.7115E 00 0.2104E 02 0.3981E 04 0.1379E 01 0.1203E 03 0.1009E 01 0.2110E 02 ~ 0.6309E 04 0.1195E C1 0.1701E 03 0.1417E 01 0.2117E 02 l 0.1000E 05 0.1018E 01 0.237BE 03 0.1965E C1 0.2128E 02 { 0.1584E 05 0.8568E 00 0.3286E 03 0.2685E 01 0.2141E 02 0.2511E 05 0.7145E 00 0.4492E 03 0.3617E 01 0.2158E 02 i ~ s i 4 0
y ^ -a. en v = = * - .W e.4Ne m=-4m p-g y g MERCitANT StilP -4: PROJECT 1-27-75 Page 2. D.3981E 0# O'. 5889 E' 40
- t. 6077E 03 O.4801E 01 0.2180E 02 0.6309E05
'. 4798E 00 0.8134E 03 0.6274E 01 0.220bE 02 0.1000E Ot- . L3888E 00 0.107BE 04 0.8069E 01 0.2241E 02 0.1584E Ot. ,1.3165E 00 0.1418E 04 0.1022E 02 0.2281E 02 s 0.2511E 06 0.2609E 00 0.1861E 04 0.1280E 02 0.232)E 02 0.3981E 06 0.2190E 00 0.2443E 04 0.1585E 02 0.238$E 02 0.6309E 06 0.1870E 00 0.3226E 04 0.1945E 02 0.2453E OR, 0.1000E 07 0.1599E 00 0.4287E 04 0.2358E 02 0.2530E 02 O.1504E 07 0.1262E 00 0.5675E 04 0.2801E 02 0.261)E 02 0.2511E 07 0.9740E-01 0.7383E 04 0.3235E 02 0.269dE02 0.3981E 07 0.7586E-01 0.9479E 04 0.3651E 02 0.2771E 02 IO 0.6309E 07 0.5879E-01 0.1205E 05 0.4041E O2 0.2844E 02 A 0.1000E 08 0.4564E-01 0.1526E 05 0.4406E 02 0.2912E 02 0.1584E 08 0.3711E-01 0.1930E 05 0.4742E 02 0.2974E 02 0.2511E 08 0.2950E-01 0.2440E 05 0.5047E 02 0.3031E 02 0.39815 08 0.2420 E-01 0.3089E 05 0.5318E 02 0.3082E 02 0.63e?E 08 0.2211E-01 0.3985E 05 0.5572E 02 0.3129E 02 I i l l
i .l n ls l ~ 7.0 RATE OF S01,tTBII,ITY OF RALON l'D1 IN WATER t ~ This subject is of concarn because a post-LOCA scenario can be postulatad ~ in which Halon 1",01 dissolves in energency cooling vster in the contai-ment and aubsequently decomposes as it passes through the reactor core. Even though the solubility of Halon in water is very low (100 ppa at 1 sta and 120*F), i.f the agent continuously decomposes in tha water and is replanir.hed, substential loss might result over a long period of time. In order to investigate this problem, two lines of study were undertaken. First, an analytical examinat. a was conducted a vSich resulted in the modeling of the problem. This modellas proved very useful j ; in understanding the sachanism of dissolution of a gas into a liquid sod in de-fining parameters of the process. The second line of study consisted of an experimental investigation of j ~ the solubility rate of Halon in water. An apparatus was f abricated that allowed water to be streamed through a 1301 atmosphere in a manner that simulated contain-L ment conditions on subscale. The experinental results were compared v3th the pre-l. dictions of the nodel. While the main purpose of the model - defining process l { parameters and showing their dependence on experimental variables - was verified, substantial empirical corrections were required in order to predict exact numerical ) { values of the solubility rate. The solubility study was found to be very worthwhilp because it showed i that the rata of dissolution of Halon by water can be quite rapid. Ilowever, since L, i the saturation concentration is so low, this primary it,sa of agent cz=not be l 'i ! important unless a secondary sourca is available for renoval of 1301. A possible secondary source in'miving decompositica in the radiation field of the reactor core must therefore be investigated. I, u lf
7.1 DESCRIPTION
OF PROBLEM AND PRELIMINARY )CDELING t A postulatea sequence of events whereby Halen 1301 mignt dissolve in i 4 emergency cooling water following a LOCA and then decoa,,ose could be as folicws: 1 A five-inch surge line is assumed to break (design basis accident). Eventually l, water at a rate of 500 gpm would be pu= ped through this line, would fall into the containment susp, and then be recirculated continuously via the low prusure in-jection system through the reactor core and out again through the breti in the five-inch line. As the water passes thtrugh the containsent vessel a:=caphere. = 2b -i
m ma l LJ I 7 it could dissolve sens quantity of Halen 1301. As it next passes through the i j reactor core, the 1301 could decompose 1 directly by the reaction ~ CF 3r + e,q + CF3 + B[ 3 which may be extremely rapid with an astimated rate constant of 2.5 x 1010 3gg,,j mole /sec. Even though the concentrations of the reactants CF Br and hydrated 3 electron, e, are..relacively low, it is still possible to calculate a large i amount of decomposition of the 1301 over a 67-day period following a LOCA. l j 1 The amount of decomposition that one calculates depends on what is l assumed about the amount of 1301 that dissolves in the water. If the rate of 1 dissolution is assumed to be infinitely fast; i.e., all the water at all times contains the saturation amount of 1301, then it can be shown that virtually all the IW1 would decompose in 67 days, if the reaction above is as, rapid as assumed. For example, the solubility of 1301 in water at 120*F and 20 psia is approximately } l 0.000136 lbs/lb. A flow rate of 500 gpm is equivalent to a total flow of 4 x 108 pounds H 0/67 days. The product of thase two quantities, 55,000 pounds, represents 2 I, the amount of 1301 that could dissolve and decompose in a 67-day period. s. The asstanttien of an infinitely fast solubility rate nust be examined. The mechanism whereby H O becomes saturated with a gas such as Halon 1301 could 2 be the following: An element of H O passes through the atmosphere of the gas. 2 p As it does, gas molecules impinge on the surface of the liquid H O and are absorbed. I L2 2 The gas molecules then diffuse into the body of the liquid, and the proces( con-r7 tinues until the F o is saturated. In the theoretical treatment discussed below. j b the diffusion of the gas in the liquid will be taken to be the rate-limiting step. No turbulent mixing is assumed. It is implied that the impingement of gas on ! l the liquid surface and its dissolution is rapid enough :o maintain a surface j isyer of H O that is saturated with Halen 1301. 2 {[ The approach to estimating the race at which 1301 vill dissolve in an element of H 0 for a diffuaien-limited case is to compute concentration profiles 3 for given assumed conditions. The profiles are computed by solvig the appropriate differential diffusion equations, from which results the expressient* f 0 p-1+ (-[ sin SjE
- e exp n=1 o
M< 7-2 i
1 ....n-..... 1 , i }t I 1 The expression is derived for a spherical geometry but is accurate for any geometry 2 {~ at low values of the dimenofonless quantity (Dt/a ). In the above equation, C i= y 'j concentration at time t. and distance r in the element of H o; C, is initial con-g , ld centration (i.e., saturation concentration at the H O surface); a_ is the dimension 2 (e.g., radius) of the element of H 0; and D is the diffusion coefficient of the 2 i y gas in H 0. 2 l Dimensionlese plots are constructed of C/C, versus r/a for various {j valuas of Dt/a (= N). Computer printed plots are shown in the accompanying {,, Figures 7.1 and 7.2 for two example cases of Naa10 and N=10 '. For the present ~ f Purposes, the area under the curves is required. This area, A, represents an l equivalent fraction of the radius,a_ that would be saturated with 1001 to the con-centration C, if all of the dissolved 1301 were confined to that fraction. In ~ Figure 7.3 is plotted the computed areas as a ' function of N. ) i We are now in a position to estimate the raount of 1301 that will dis-solve in H O as a fun tion of time. We vill first consider the case of the design 2 basis accident since this is the only case for which the flow rate and df **= ton of the water element can be specified (500 gpm and 5 inches diameter, respectively). ~ The corresponding value of N is taken to be 2.48 x 10 This is derived from a i value of D=10'
- m /sec (an estimate based on similar systeca**); and from a time perimd of 1 second, assumed to be the length of time required for water to fall s
in the containment. For N=2.48 x 10, the corresponding A is 5.4 x 10 (from Figure 7.3). ~ c. { This A-value represents the equivalent fraction of the radius of the 5-inch dia-meter water stream that would be saturated with 1301 in a 1-second period. The e,i corresponding fraction of the total (cylindrical) water stream that becomes t saturated is therefore 1.08 x 10 3, or 0.108 percent. This latter quantity is t j ] referred to as the theoretical F-factor. Accordingly, the assumption of infinitely { '- fast solubility rate and complete saturation of the water stream as it passes l .I Into the reactor core would not be nearly satisfied on the basis of this theoreti-cal model for the condition of the design basis accident. Instead of the previously i } 'H.S. Carslaw and J.C. Jaeger, " Conduction of Rest in Solids," 0xford University Press. 1947. ( !nternational Critical Tabig, Volume 5, p. 63 s. 125c 7-3
r ~ r _.. ~ __ ___.._______F t [_ t C C' e I~; i .i. C/CD VS R/9 } I o lilNFl ? 197E ~ 3 i I j e.ere ::ce.N i o t e i o. \\ O t.: ci-t a P Y i N o 7. e i c; g I U i N u ~ o g-1 c.. M "O3.70 93.73 9'3.77
- 99. Ta l
- 39. r25
- 99. :53
^ 99.92 9'9.96 100.00 R/4 x10^* Figure 7.1. Computer P,rinted Concentration Profile for Diffusion-Limited 4 Model of flalon 1301 Di>.olution in H O (N = 10 ). 2 t F M 4 _.._____.__.___,m m
m----, a ss. .am .----s s_,4 x a s A l ,1 j i 2 4 e 43 To de $i: o l O n a. f. O F f i-1 (O $a M co o, Mao 7 O* l 'ra gE G o "h K
- $ N
~ qZ \\ i. t.E -c 50 3 E m m
- 2 l
~ i.8 ec P $Q o-UC21 v s. - c ,P k b. r. ,, Z e. s "n. 2g i D (* C N $3 T u2 j, n C N a to f i N N (C C ,ji 6 (O C
- \\
'n C L U M dm q ll U 1 y t~ ' j O A i 0:l ' i CD'G 60 oh 0 02'O C C 0' 7 03/.3 7-5 127< esguaeee - ~-g.
- - - -.. ~. l ,~- y. c._ a__. L_ 1 i 10*I A / / / / / ~ 10-2 ~ 1 f.r 4.e e w EC / u / P 1 / 10-3 O C-. ~ /__-._.. j r ~ 4 ._~ 4 1D 107 10 6 10-5 104 10-3 10-2 i Figure 7.3. Plot of the Guantity N versus Area A of Concentration I Profile Curves in Solubility Rate Model. i 6
x - ~r N i I i.a computed 55,0C0 pounda of 1301 that could dissolve and decompose in a 67-day period, the quantity would be only 59 pounds. Assumir,g a total quantity of l Halon of 20,000 pounds, this loss would only represent 0.30 percent. j] The above example calculations were chosen te illustrate the theoreti-i cal extent of 1201 decomposition that can te expected for the only case in which operational cceditiona can be specified. The extent of aalubility for any other assumed conditicca is readily computed from an estimated value for N and the plot ,) shown in Figure 7.3. However, the flow rate of water aunt be specified. yor J example, if one assumes that water vill discharge as 1-inch diameter spherical l droplets instead of a 5-inch diameter cylindrical strean, and also chooses 5 i, l] seconds rather than one second as the time period for exposure to 1301, then the l corresponding N-ealue is 3.1 ~r 10 and the A-value is 6.2 x 10 The calculated -5 ~ f ] fraction of each droplet that vill become saturated with 1301 (F-f actor) is i t 0 0186, or 1.S6 percent. If the DBA gallonage of 500 g; m is assumed, then 9.65 percent of the Ealon would dissolve in a 67-day pericd. On the other hand, if .1 the 500 gpm figure is reduced to a lower rate, say, for exa=ple, ty raticing the 2 ({ areas < corresponding to 1-inch and 5-inch diameters (i.e. 1 /5 x 500 spn = 20 gpm), then the a:nount of 1331 dissolution becomes 0.13 percent. } The a.ove two examples serve to illustrate a diffusion-linited solu-I bility-rato nedel. There atu, of course, an infinite na=ber of such examples ) <j that one could choose. However, as will be seen later, large e=pirical coef fi-cients are required in order to predict solubility rate frra the model. The main. reason is that a water strean, as it f alls through an at:nosphere, tends to strongly break apart. Thus, the assumption of a solid stream is not fulfilled. l There is also stistantial turbulence within the water stream. The modal has to I be corrected to account for these effects by adjusting the N-value, which corre-i sponds to increasing the effective diff usion coef ficient ed/or decreasing the ef fective radius a_ of the water columni or droplet. Because the magnitude of the i correction f acters is large, the usefulness of the model in predicting exact nuwrical values of the soltbility rate is linited. Ecwever, the modal serves to define the solthility rate techanism and to ehuv the effect and dependence of certain para =eters. ne question of the rate of solubility can only be defini-l tively answered by experinectal t ening. Such a study was ccnducted and is des-cribed in the fellowing secti: a. h 1. ( 129< l
- 1. 7
w_ ~ 6 o 7.2 C:PERIMDrfAL APPARATUS AND METliOD OF GIDt1CA1. MA1.YSIS 7 In order to study experimentally the rate of solubility of 1301 in water an apparatus was constructed which minulated the containment configuration on a smaller scale. A achamatic diagram of the apparatus is shown in Figure 7.4. he test section consieged of a Pieriglas tank 10 feet in height by 17-3/4 inchas g L in diameter. He f ank was equipped with a copper inlet flow tube located at the top through which water was continuously circulated during tests. A water stream t issued from the tube and fell approximately seven feet to the bottom through a f( Italon 1301 atmosphere which was maintained in the tank. Parameters of the study which could be controlled were the flow' rate of the water, the size of the water stream, the concentration of 1301, and the system temperature. The working of the apparatus in detailed below. he tank was filled with 133 pounds (16 gallons) of water prior to each test and the water tag erature was thsn set by means of the electric immer-sion heaters. The top of the water surface was covered by a flat plate no that only the flowing stream was exposed to 1301 during a test. D e flat plato con-e I tained a "spinah suppressor" which consisted of a honeycombed metal cylindur about 5-inches in height and 5-inches in diameter. This suppressor very eff ee- ] tively captured the water stream and returced it to the reservoir without creating splash or causing bubbling below the surface of the water. The inlet 1 ) flow tube was a copper tube 2 cm inside dia:neter by about two feet in length, a highly polished on the inside. Water was circulated by means of a pump through i "I a rotamater-type flovmeter which measured flow rates of up to 8 gpm. Italon 1301 was admitted to the Plazigiss tank from a supply cylinder. ' 1 The. concentration of 1301 in the tank was controlled by using a calibrated gas 1 measuring cylinder to set the amount of Halon to be admitted. Either 0.5 or 1.0 atm of 1301 was used in the Plexiglas tr.nk for tests. To attain these concentra-tions, a known pressure of 1301 was first admitted to the calibrated gas measuring i cylinder, and then, this 1301 was discharged into the tank. The agent entered thi i l tank near the bottom and as it filled the tank, air was displaced at the top. l This method worked very well, and since the density of 1301 is five tirre greater ) than that of air a sharp interf ace between the two gasca could be seen as the 1301 was being admitted to the tank. When 0.5 atm 1301 was employed, a gas mix-ing fan was used to mix the 1301 and air in the Plexiglas tank. i 130< [ 7-8 l
+ 1
- 1
~~ ^ ~ ~ ~ ~ ~ i...... i i___i C u_. I L.. C [~ L.. _ tT-F- i + 5 t r 'f ALUMINUM PLATE I GAS MIXING FAN Na i PIPE ~ j CIRCULATING WATER INLET FLOW TUBE WATER STREAM / CAllBRATED HALON l GA5 MEASURING TANK 17 3/4 IN. DIA.X 10 FT [ PLEXIGLASS TANK i PRESSURE l CIRCULATING WATER GAGE FLOW METER N Y FLOAT AND SPLASH i g l SUPRE550R }A - ~/ is o#[ WATER a LIQUlFIED HALON 1301 l l _ f-SUPPLY TANK ELECTRIC HEATERS DRAIN 4-Q-
- 'I #
et 4 i WATER CIRCULATING / h gl THERMOCOUPLE PUMP s ( 3 HALON FLOW METER WATER SAMPLING INLET COCK Figurs 7.4. Schematic Diagram cf Apparatus Used for Study of Solubility Rate of Halon 1301 in Water P G
y ~ ~ he ' f g. ] Tests were conducted by flowing water through the 1301 atmosphere for given periods of time. Periodically throughout each test, samples of the water were taken through a sampling cock in the water line,and the concentration of d< 1301 present in the water was determined by chanical analysis. A method of chemi-l cal analysis was developed specifically for this project and is described bel:,v. The water samples emtained between zero and 300 ppm of dissolved Italon 1 1301. Once a sample was taken it was fotnad to be crucial not to expose the water to air or to ullage volume of the sample bottles even for a short time be-cause 1301 gas would escape from the water. Samples were analyred using a gaa I chromatograph with an electron capture detector and a silica gel column. Initially, portions of the water samples containing 1301 were injected directly into the chromatograph. Ilowever, water hold-up on the silica gel column became an inter-l 14 ference. Accordingly, it was necessary to extract 1301 from the 110 prior to 2 l injecting into the chromatograph. I ) ne water samples were extracted with heptane solvent. Virtually all { of the 1301 entered the heptane phase since its solubility in heptene is very I great (6.3 percent by we16ht) relative to that of !! 0 (0.03 percent by weight). 2 Portions of the heptana solutions were injected into the chromatograph. The method was calibrated by dissolving known quantities of 1301 in heptane to prepare s tandards. The standards were checked in two other ways. First, natut ated solu-l,l l L tions of 1301 in water were prepared by shaking 110 in the presence of the gas, 2 i These were then analyzed in the sa:an way em test samples. Secondly, gas standards t }', of 1301 in air were prepared and injected directly into the gas chromatograph. ] initially, test results were very erratic until familiarity with the { appara;;us and experience with the method of chemical analysip was gained. There- \\ af ter, few problems were encountered. Tabalated below are data of the maximum ( [ amounts of 1301 found in each test perfomed, plus data from several separately prepared saturated solutions. nese should represent the saturated concentrations . F j at the temperatures listed. They can be empered with data reported by DuPont. ~ 1 l 132< 1 T I b i 7-10
..esa.nnum p. w Analysis of Saturated Solutions of Halon 1301 in Water (ppm by Veight) 1 atm, 77'F
- 0. 5 atm. 77
- F 1 a ts, 120
- F 0.5 atm. 120'1' 272 103 95 55 288 e 143 91
] 306 185 81 268 167 83 322 115 E]' 150 308 27 (-1182) 93 362 7 10 (=10%) L, 312 ] Average 302 "2 122 (=17%)* I The DuPont Product Bulletin (B-29C) on Halon 1301 gives the following ssturation n. solubility data: dj_ Temperature Pressure (*F) ppe1301inHg ] 1 atm 77 300 iJ 0.5 atm 77 4 150 1 atm 120 N 100 j 0.5 atm 120 % 50 j It is seen that the agreement with the DuPent data is excellent, especially con-sidering the fact that the analysis of a solution for er.all amounts of dissolved ] gas is a very difficult analytical chemistry problem. For reference purposes, the solubility of a few other common gases in water is tabulated below: Handbook Data on Solubility of Gaaes (in in) in Vater et I atm Pressure u T(' F) 32' 77
- 120' f*
CO 3350 1450 760 2 ~ N 29 l8 12 2 { 0 69 39 27 2 2-Air 29 17 10 (35 0.) (34: 0) (33: 0,) 4 de { 133< L11 L.
r J 7-All gages, ruters, thermocouples, etc., used in this task of the program e were e=1*hrated by ARC Quality Assurance procedures, traceable to NES. ~l j
- 7. 3 EIPERI.Vl2 ITAL TEST RESULTS
} In this section of the report we vill present the experimental results L in the form of tables and graphs, and reserve analysis of the results to the following section. Table 7.1 sisnaarites the test conditioca for 18 experiments ,j en the solubility rate of 1301 in H 0. Essentially two 1301 pressure levels 2 (0.5 and 1.0 sta) and evo temperatures (76' and 120'F) were used. Tempe ratures control was a problem until tast utssbar 6. Water flow rate ranged up to 8 gps. ~ The effect of the splash supptessor was tested by removing it during several tests. Tests with zero water flow rate vera conducted with the top surface of the H O 2 J . in the test tank exposed to a Halon atmosphere. No water strema was flowing in these two tests (numbers 13 and 14), however, the 16-ga11rm body of water had to be stirred just prior to sampling each time to insure unifermity of concentration. [ { The stress of H O that issued from the flow tube at the top of the tank 2 I would 'resain as a aclid colum of water for some distance but would eventually break into stresslets prior to reaching the splash suppressor. The f all dis tance prior to break-up depended on the concentration of 1301 in the atzsphere, the 1 i temperat=.e, and the water flow rate. As all three of these increased, the dis-i tance to the break-up point decreased. At 77'F,1 gpa water flow and no 1301 present, a solid stream could be maintained almost to the bottom of the tank. I At-120*T 1 atm of 1301 and 8 gpm water flow, the solid strema vould break af ter ~ about a two-foot f all distance. Referring again to Table 7.1, the colu:m labeled "H O Col:mn Diameter" represents measured diameters of the water columns at their 2 mid points in the tank for various water flow rates at 77'?. From these values, a fall distance of 206 cm, and the gallonages, a total fall time for an element of H O can be estimated. These are listed below: l 2 t I H.,0 Flow H O Column Stress /es__j Tall 2 Rafe (kon) Dicerer (cm) Velocity inec/ Time (see) 1 0.37 587 0.35 2 0.49 669 0.31 3 0.61 647 0.32 8 1.05 583 0.35 b 134< 7-12
' ' ~ e ~ J- ) J Table 7.1 Summary of Test Conditions for Solubility Rate Study iJ Test ,1301 Flow Rate Temperature H O Column Splash 2 No. (ste) Water (gpm) (*F) Diameter (em) Suppressor -} ~ ~# 13 1.0 77 9 1.0 1 76 0.37 Yes ~ J 12 1.0 1 78 0.37 . No 3 0.5 2 68 0.49 Yes 4 0.5 2 72 0.49 Yes 5 1.0 2 74 0.49 Yes 6 1.0 3 73 0.61 Yes .J 11 0.5 3 76 0.61 No 7 1.0 8 76 1.05 Yes 8 0.5 8 76 1.05 Yes 10 0.5 8 76 1.05 Yes 14 1.0 0 120 16 1.0 1 120 0.37 No 19 1.0 1 120 0.37 Yes 17 1.0 2 120 0.49 Yes 18 0.5 2 120 0.49 Yes 15 1.0 3 L20 0.61 No 20 1.0 8 120 1.05 Yes .J e I 135< 7-13
yg ~ l The fall time, whi:h represents the cine that an element of H O has 2 {~ availabla to absorb 1301 dtring its residence 1:n the atmosphere, was relatively l t constant in all tests. As a point of interest, a velocity can be cale-lated
- f for acceleration due togravity alone. This is given by v = /Sg/2, vbere g is 2
32;2 ft/sec and S is the fall distanea, takam to be apprcximately 10 feet. a V is therefore 387==/:c.c. Cince this value is insa than those tahulated above, it is apparent that the water pump is adding =* rity to the H O stream, 2 The raw data for each of the testa listad in Tabla 7.1 are r.ot pre-a sented in this report in order to conserve spea. The data are svailable from Todd or Atlantic Ramearch Corporation. A ntaber af the tests have been selected and the data plotted in the accompanying Figures 7.5 through 7.10. "ihe tests selected to be plotted were among the most rr,+%e4hle, but were nainly chosen ~' because they illustrate points for discussion is the following section of this report. We vill briafly describe the results giwan in Figuas 7.5 through 7.10. The plots in Figure 7.5 illustrate the rare at st.ich Raion 1301 is f absorbed by H O at 77'F and 1 atm pressure for ufr.ree H O flow rates of 1. 3 and 2 2 { 8 gym. Fron Figure 7.5 it can be calculated ria: the rata of solchility of l 1301 into H O was 3.3, 8.0 and 33.3 ppm / min a: ". 3 and S rps flow rata, respec-2 tively. ~hese rates refer to the Halon absorte.1 ty 16 gallons (133 pounds) of f H O dur.ng the tests. The dashed lines on the fir:re vill be di ::ssad later. 2 The data plotted in Figure 7.6 shov =be solubility rate for a quiescent l test with no water flow, at 120'F and 1 atta 13*.~ In Figura 7.7 the solubility I rates at 76*F and 8 gym are compared for two Merant partial pre.ssures of f 1301. These results also illustrate the seltii'iry equilibrium crecer.tration l 1evels at this temperature for 0.5 and 1.0 a:s 13*1. Figrre 7.8 shows the l equilibrius concentrations also at 0.5 and 1.0 a=2 Halon bct at a teeperature l of 120*F. Figure 7.9 illustrates the increased rata of absorption of 1301 when no splash suppressor was employed. Altictugh the data of Test 9 shev appreciable scatter, it is apparent that the er:".siility rata in our tests was about twice as great without the splash suppread.:r. In Figure 7.10 the data of Test 6 a+= -lotted in order to compare the visental results with the theoretical model. *his figure vill be disc =s s.ed beJov. 136< 14
1 i i J 'l l J 350 I l TEST 7 1.0 atm 1301 8 gpm H O 76*F 2 300 M 's-J ,( r m ^ l O ~ 250 3 gpm H O ) .) TEST G J l (' / 2 200 l ( TEST 9 .) ~) -lE 1 gpm H O 2 l a i 150 I [.) TEST 7 THEORY /p/, ~~' [-] f top D/ / l / TESTC / THEORY s-I s i 50 / / TEST 9 ',,/ / THEORY s s. i j 0 30 60 90 120 150 180 210 TIME (min) { l i . i Figure 7.5. Pfot Showing Effect on H.O Flow Rate on Solubljity nate of Halon 2 1001 In H 0,(0.value forTheoretical Lines is 10* cm j,,,y, 7-15 137<
M I~ !~ .~1 L 3 C C C1 i.. _; m pg m pg g g g pg l 6 k l f 125 I 1.0 atm 1301 i 120'F NO H O FLOW 2 100 o / TEST 14 75 Y m 1 /, ,s' e y a. P ,L 50 / C) w s' CC i THEO Y 2 = 10' cm j,,, j 25 l THEO Y 2 D = 10' cm /sec ---h---- 0 0 250 500 750 1000 1250 1500 1750 2000 TIME (min) l Figure 7.6. Rate of Solubility of Halon 1301 in H O During Quiewent Test. 2 L 4 r l' e
O' .. - _.....,. _,... ~.. -........... _... -.. a 1 J 350 g 76'F 8 gpm Hgo 300 4 7 TEST 7 1.0 atm 1301 ~ l { ' 250 200 i g TEST 10 .J p I P.5 atm 1301 0 ~ /F kJ 4> OH / /p/,(TEST 7 4 THEORY 100 / I /,/ l s TEST 10 e'p THEORY q / 'p f . 2' 50 p p ',s - / s /s,s' ~ / 0 l',s' / 0 30 60 90. 120 150 100 210 TIME (min) F; 1J .E I Figure 7.7. Plot Showing Effect of Halon 1301 Partial Preuure on Solubility Rate and on Equilibrium Concentration. 7-17 ?. 138-
,l e 1!t}} i.1 i i. = ti 4 i. E i.. .e 1 F 0* 1 2 l e b = 9 i 0 ~ 8 1 5 C 1 0 3 1 0 5 n i_ 1 o l la .e Hr i u ~ f s o se C O 1 nr 0 oP i 0 3 t m 1 a 3 8 0 r t 1 m 2 n. t 5 1 m Tt e 1 n. 1 a Tt S ) c a I S E 5 in nod E0 T0 m C n T1 ( a E m5 O M u 0 i I r 0 ibd T n a 9 il a uqF E *0 g 2 n 1 i A V wt o a h0 N S 2 0 t 8 / lP In 6 o1 8 7 e ru ig 0 F N 3 I O C r 0 0 g 5 o C 0 2 1 5C d* M ~ ne l H df.
O' T A. 350' I 33 .x ~~ 300 L \\ LTEST 12 250 ^ J ] i t1 200 O \\ oU LTEST9 2 4 150 (.) o+ (.) l00 Ol C! 1.0 atm 1301 8 78'F 1. I spm 50 I.. 9 T. 0 0 30 60 90 120 150 180 210 ~ TIME (min) l l 1 I I v t d' j Figure 7.9. Effect cf Splash Suppressor on Solubil!ty Rate of Halon 1301in H OI" I*" ^PP''" 3 2 7-19 0 141< y,w m me ~ ~ '-
s I 300 TUT 8 3 L 250 .I THEORY 1 D=j.2X1d / q. cm /sec d 200 / 3' ~ / / I l50 I / .,j d / E / / m, 'i. / U 100 / // [* [ f THportY D=10 5 / cm'im 50 / / l' -], I. ~ "**( lj; 0 '"**~ ~ f 0 10 20 30 40 50 60 70 } TIME (min) L 1 t- $b 6 .t - L. } ."Igure 7.10. Compana of Data from Test 6 (3 grm H.0: ?6*F1 ~ ~y 1.0 stm 1301) =tth Theory. i ' 20 1 11&_
O. l 'I d 7.4 DISCUSSION OF RESULTS The first point to be considered concerns the accuracy of the theoreti'- cal model in predicting exset numerical values of the solt.bility rates. Computed 11 ] L. theoretical curves are plotted on Figures 7.5, 7.6, 7.7 and 7.10 for the respec-tive conditions of those tests. This was done by the method described at the beginning of this section. First, en N-value is estimated. To do this a diffusion ~ coefficient of 10 cm /see was used; a residence time of 0.31 to 0.35 see (see f~' Table,7.2)r and a dimension a corresponding to the measured H O column radius 2 given in 'lable 7.1 (dismeters are listed in the table). For the tests of Figure 7.5 l t {' the following values were used: L. Test N-Value A-Value F-Factor l 9 1.02 X 10 ' 1.15 1 10 0.023 -2 ~ -5 ~3 6 3.44 X 10 6.7 I 10 0.013 -5 ~3 j, 7 1.27 K 10 4.0 X 10 0.0080 L It will be recalled that the F-factor represents the fraction of satura-t- tion t, bat the water colunc achieves on one pass through the 1301 atmosphere. It is obtained from the A-value by strictly geometric considerations. There ,{ were 16 gallons of H O used in each test. This quantity divided by the gym 2 rate gives the time in minutes that it takes for all the B;0 to be exposed once to the 1301 atmosphere. That is.for all the water to reach the fraction of saturation given by the F-factor. These fractions are converted to ppm and plotted as the theoretical curves in the figures. The theoretical rates of i. saturation are linear because it is assumed that each time that an element of water is exposed to 1301 during a test it increases by a constant fraction. More correctly, the 1301 content of an element of water will increase by a smaller fraction with each pass. The fraction will decrease in proportion to the amount of unsaturation remaining. However, for present discussion purposes l. the linear rates are adequate. The theoretical line in Figure 7.61a not linear because there was no water flow and the model was applied with no racirculation. In Figures 7.5, 7.6 and 7.7 it is appe. rent that experimental rates greatly exceed theoretical rates. It is possible to curve-fit the data using "ef fective" N-values. 'lhis would correspond to increasing the diffusion co-efficient D and/or decreasing the stres:n radius a,. Physically, these in turn correspend to stream turbulence and breakup of the water colu::n. For illustra-tion purposes we have chosen to use " effective" values for diffusion coefficients 7-21 143< t L
e IJ only. This is done in Figures 7.6 and 7.10 where it is seen that diffusion coef-ficient values of two to three orders of magnitude greater than the value of the molecular coefficient (10-5,2/sec) must be used to fit the data. Tabulated be-3d low are effective diffusion coefficients obtained by curve-fitting the experi-mental data, for the gallonage rates of Figure 7.5 (76*F and 1 atm 1301): h CPM Effective D (es'/see) , u -3 1 0.6 x 10 -3 3 1.2 x 80 8 6.3 'i i ~ Even though the D-values are increasing at a decreasing rate, it is not apparent i; how to scale such data, say, to DBA conditions. Moreover, the experimental rata data at 120*F show enough scatter that it is not possible to distinguish in-g creasing solubility rates at increasing !! 0 flow at this tes:perature. It is 2 [, concluded that the theoretical model is probably not useful in predicting exact ntanerical values of the solubility rate. Accordingly, we have chosen a more empirical way of snalyzing the data which is presented below. L It is possible to calculate actual F-factors from the experiaantal i rate data. We data of test 6 in Figure 7.10 can be used as an exseple. ne
- s..
first 50 percent of saturation (150 ppm) was achieved in just about 19 minutes which corresponds to a solubility rate of 8 ppa / sin. This in turn corresponda ~3 to an absorption rata of 1.07 x 10 lb/ min of 1331 for the total 16 gallons of water being circulated. The flow rate during the test was 3 spa, so the time for all of the water to be circulated once during a test is 5.33 minutes. Hence,16 gallone water absorbed 5.67 x 10' pounds of 1301 per pasa (for these test conditions of one-third second expesure time,1 atm 1301 and ~ 76'F). The 16 gallons of water can dissolve a tetal of 4.00 x 10 pounds of Halon 1301 at complete saturation of 300 ppe. There. fore, the fraction of saturation achieved per pass of the water through the Halen atmosphere is 0.14 { (or 14 percent of saturation). This represents an experimental F-factor for up to the first 50 percent of saturation. m. Experimental F-factors have been calculatad for all the tasta per-- formed in the study sad are collected in Table 7.2. nere is no correlation between the F-factors and the water flow rates. 3.. wever, there are other trends 1. 144< 1 7-22
= Tab 1's 7.2 Experimental F-Factor in Solubility Rate Study Test 1301 Solubility Rate S plas.h Temperature H O Flow F-2 No. _(atm) _ (pps/ min) Suppretesor (*F) (gpm) Factor 3 0.5 2.5 Yes 68 2 0.07 l 4 0.5 5.0 Yes 72 2 0.13 t, 10 0.5 12.5 Yes 76 8 0.08 ~ 11 0.5 16.7 No 76 3
- 0. 30 5
1.0 9.4 Yes 74 2 0.25 k-6 1.0 8.0 Yes 73 3 0.14 7 1.0 33.3 Yes 76 8 0.22 9 1.0 3.3 Yes 76 1 0.18 12 1.0 13.6 No 78 1 0.73 I.. 18 0.5 1.4 Yes 120 2
- 0. 11 i
L 15 1.0 8.3 No 120 3 0.44 f 16 1.0 4.2 Uo 120 1 0.67 \\ 17 1.0 2.4 Yes 120 2 0.19 19 1.0 2.4 Yes 120 1 0.38 20 1.0 4.2 Yes 12 0 8 0.08 Ie- _j 'l } t i 145c l 7-13
1 ] Table 7.3 Susanary of Average F-Factors for various l 'A Conditions in Solubility Rate Study 'l l l r 0.5 at 1301 1.0 atm 1301 0.5 at 1301 1.0 atm 1301 76*F 76*F 120*F 120
- F With Splash 0.09 0.20 0.11 0.22 Suppressor I
Without Splash 0.56 0.30
- 0. 7.,
9 Suppressor l J F-Factor: 3x greater without splash arrester 1x higher at 1 atm than 0.5 atz Approximately sama at 76*F and 120*F + J (Dwell time of 1301 in presence of water %.33 seconds in all cases) A fl L) l J Fia e b4 9 i4 146< l-7-24
It' ' a l I which are summarized in Table 7.3. It is seen that the ava age F-factor is apprordna ely 0.2 at 1 sta 1301, independent of temperature'. At 0.5 atm 1301 the factor is one-half of this value, also independant of temperature within the precisias of the data. E11mination of the splash suppressor produced about a three-f:Id increase in the F-factor under all conditions. e [ The F-factors listed in Table 7.3 are very useful because they can be used to esHware solubility rates undar othat conditions including those that could be poetnisted in a containment. As an example, an estimate for DBA con-a ditions ema be made. The dependence of the F-factor on various parameters is given belar tesether vith the rational for the choices. ? I Parame te r F-Factor *ependenev r* T v atars Independant - Based on arperimental results b Pressure Ealon 1301 Linear - Based on experimental results H O now Eace Independent -- Based on experimental results 2 f Square Root - Assumption based on model m. i ~ The DEA ccnditions vill be assumed to be 500 gpm water flow; 20 psia pressure 1301; IO*F; one-half second fall (residence) tine of H O in 1301 atmosphers. 2 The F-factor of 0.2 is therefore corrected to: m. 0.2 X 20/14.7 x (0.5/0.33 0.33 = The vatar flow rate is 500 gym (4169 pounds) and the saturation concentration of 1301 at 23*F and 20 psia would be 136 ppm. Eence, the solubility rate of 1301 into water u= der these conditions is computed to be 0.187 lbs/ min. In a 67-day ] period (H.490 minutes) this corresponds to a potential removal of 18,042 pounds of Haloc l'a31. I I' It is important to note that this last value for the quantity of Halon i* 1301 that cocid be lost necessarily requires a secondary source for removal of y-1301 fr:xa tim water. The solubility in water alone, which cos.1d be considered the prinary removal source, would not account for much loss of agent. Such a secondary so=ree has earlier been postulated as 1301 decon: position while the [ energen y c: cling water passes through the reacter core. Without such a loss me ch =- f w fer 1301, its depletion in the containment due to dissolution in H,0 3 ~ vould be na g.11giblo. Assuming, for axa=ple,11,000 f t of emergency cooling 147< 4, 7-25
Q-- IL water, tha total amount of 1301 that would be required for complete saturation g (at 120*F and 20 psia) is only 92 pounds. This small amount of water required ( to simply saturate the cooling water is less than one-half of one percent and i would represent a negligible loss. a Thus, while the conclusion of this study is that the rate of Halon 1301 l ' f. solubility,into water can be rapid, the question of how much Halon would be lost from a containment following a LOCA is still dependent on whether the agent would 1 q J subsequently decompose. To answer this question, a study of the radiolytic de-composition of 1301 in aqueous solution will be required. I 9 U 1* e 148< j. i 7.74 I l l
\\ 1 r l I v 8.0 I.*tELIMINARY SYSTTM ANALYSIS a 8.1 SYSTEM CONFIGURATION J In this section we present a preliminary configuration of a Raion 1301 suppression system for a comenh==t application, and describe the procedures and principlas to be used to determine the quantity of Halon 1301 required and the rate at which the agent has to be injected to insure protection at all times follohg a LOCA. As win be seen later in the next section, of the order of I 20,000 pounds of Halon will be requ1 red for the containment volume and conditions l that will be assumed for sample calenlation purposes. E-nr=, we will use this quantity as a working figure. Were are a ntasber of Mn-n factors which win not be addressed in this " ~ ~ report. Fo'r example, it la not yet' decided whether,' following a LOCA and af ter hydagen concentration has reached the lower fissmability limit (taken as 4 percent g by volume), the total quantity of Halon (20,000 pounds) win be injected at one time (i.e., over about a ena-day period), or whether a programmed injection schedul' will be followed over an ="*aded period of time. At the present tima e instrumentation is being evaluated to chemically monitor; the gases in the contain-1 l ment, thus the programmad injection option might be available. Eovaver, for pur-g ,s g poses of this report complete injection starting at 4 percent H will be assumed. 2 c. Other undecided questions are: lihether injection win be initiated manually or f autometically or by both means; sad whetter 1301 storage vessels will be located f9 inside or outside of the reactor ccapartment (they win not be in the containment). l Shown in Figure 8.1 is a proposed configuration for a Eslon 1301 system. ~ E, "* This is based on a Babcock and Wilcox design and is inter.ded to be preliminary only. f Some faatures of the diagram to note are the following. Redundancy is incorporated f. into the valving, the itema and the injection nozzles. Five 1301 storage vessels t i
- C are shown.
{. This would represent one acre than required. If 20,000 pounds of Ealon are required for inerting, then 25,000 pounds would be stored in five vessels. 5 The Halen would be stored as a liquified gas. A conservatively high storage tem-I perature of 150'F win be assumed, although the actual storage temperature is t expected to be lower than this. At 150'y the density of 1301 is 60.7 lb/cu f t (OuPest Product Bulletin 5-:9C). A single storage bottle of 82.4 ft would be required to contain 5000 p;t:=ds of Sales. *his correspe=ds to a sphere of 5.4 feet O 149c 9* S-1
.a i 1 ~, { - -{ p* y p, gj Q Hden 1311 l@ ties I b' [M ] m2 g ;;:ag-a :12 ;, w e r_ m
- le
- e
- e
- e a
a r a Air Test Connectoe O o A ^ p p& p p 1 e For Hden Nea$es h coon costs coots coore , cects sI en e, Metas:
- 1. The syntsat is AsasE tit. Class 2.quettty youp cissifkation S.
- 2. The sycans wel conspo with the seguireneemts of Nf?A 12A.
Centainment Wet Wess r Osbcock and Waces Preliminary Design Omiy Rotemece Deswing 5ydk Dwg. Na, PSAR Fig.No.
- 1. P & 10 syvmbolo 174247 E 8-I
- 3. Router ceenpertament 20ll20 E S*IO veshtime 1 Ceepos.es.t Coeling 357858 F 9'3 Figure 8.1. Post-1.OCA Combustible Gas Control System.
e
y ~ -- --.-. _.-. -...... -. _....... - u. 'me in diameter. (The equivalent diamatar of a single sphere to contain 20,000 pounds is 8.6 feet.) a The vapor pressure of 1301 at 150*F fa 559.2 psia. A wall thickness for the storage vessels of approximately 0.65 inches is estimated by assuming material with an allavable strees value of 15,000 psi ed a storage pressure of 600 psia. A nitrogen bottle containing sufficient gas to maintain a 360 psia delivery pressure t. is tha 1301 vessels would accompany each storage sphere. If a conservatively low temperature of 75'F is assumed for 1301 storage (vapor pressure 229 psia), then nitrogen bottles of 1.84 cubic fest volume at 6000 psia pressure would be required (ig inch diameter bottles). The foregoing calculations are meant te serve only as an example to i ] Provide approximate estimates of the size and requirements of a 1301 suppression Similar computations would be made fer an actual application after all eys tam. system specifications are defined. A final design would comply with ASME Boiler j and Pressure Vecsel Code, Sectinn III, Class 2 Components. The specifications in NFPA Standard 12A would also be used for guiha. j 8.2 SAMPLF. CALCULATIONS OF HALON 1301 W!C }d In order to determine the Halon 1301 requirement in a containment vessel following a LOCA a number of system conditions must be specified. For illustra-7 [, tion purposes the following conditions v111.be asstanad (mostly based on preliminary information supplied by Babcock and Wilcox): ) [, Temperature: 120*F, assumed constar.t Initial Pressure 5 psia total - 1.7 peia stesa at 120*F and 3.3 psia air f' Containment U11 age Volume (Dry Well): 35,000 ft 3 3 3 Wet Well Volume: 17,500 ft - 7,200 ft ullage displaced air into dry well raining dry well initial pressure to 5.7 psia (4.0 psia air) Containment Leak Rates dP/dt = 0.002 P (day-1) Metal-Water Ranction: 271.45 g molas I; (214.9 SCF) produced immediately by zirconium e The rate of hydrogen generation in the containment must be known. In this exacrple, the rate based on the !CS Corporation computer code, presented in Appendix A of Section 6.2 of this report, is = sed. The assumptions required by 151< i 8-3
i 1 1 e i USNRC Regulatory Guide 1.7 vere used, and other assumed conditions are listed y a with the NuS tabulation. No sitconium-v.ter reaction was assumed in the NcS l eticulations. Ite tabulated data for H2 generation fr m the NUS code are plotted in Figure 8.2 for a 67-day period. The rate of 02 generation is taken to be one-half of the rate of H 8***#*'I""* 2 The above information allows the calculation of the gas composition in the contmimwt t as a function of time. This is shown in Figure 8.3 in the form of partial pressures for N, 0, H, and the volume ratios of H /0 and N /0
- j 2
2 2 2 2 2 2 As an example, at 0.2511 x 10 seconds (2.9 days) there would be 1861 SCF of radiolytic hydrogen plus 215 SCF of zirconium hydrogen. In 35,000 cubic foot volume at 120*F this corresponds to a H2 pressure f 1.03 psia. The exact values of the other gases at 2.9 days are: I l P 103 Psia (18.82) H2 P
- 1. O psia
= O2 P 3.16 psia N 2 PTotal = 5.49 psia (onitting steam and Halen 1301) H /0
- 0. 79 2 2
) N /0 2*03 2 2 In order to detemine the amount of Halen 1301 required at any time, { the results of Figure 3.11 are used together with the N /0 ratios and tha total 2 2 pressure including the 1301. Strictly, this requires en freration, but in fact since the pressure effect found in the explosion-limit measurements was so small, it is simpler (and conservative) to assume a total pressure of 3 ats. Then, for the above example, at 2.9 days vf th the N /0 rati = 2.43, a peak percentage of 2 2 Halen 1301 of 42 percent is read from Figure 3.11. This represents 3.98 psia of 3 e. Halen 1301, which in 35,000 f t at 120*F corresponds to 3374 powls of the sgent (DuPont Product Bulletin 3-29C). In a similar manner the weight requirement of 1301 throughout a 67-day g-post-LOCA period can be calculated. Results of the calculatten are plotted in Figure S.4 The hydrogen concentration does not reach 4.0 percent until approxi-b mately 1.0 hours, and by the end of the 67-day period a total of 19,200 pounds (22.65 psia) of 1301 are required. i 9-4
- ~+ - ~.,-r.. - ~p--.- ru~ -- - f r mimU.v3,.w u --sa;.; sca ;---CAa r1 as/5:CE41.2 bh L;M,U . -- a t-i sw M >~a H H M P=4 >=4 and Emed tune aus aume soit ene sets t=1 4 12 f i 1 10 t l. 8 / 5 n O I i X 1
- c. 6 l
T r a I p u l C11 o" C: A I t 4 il I 1 i l. !l 2 i 1 a 0 { 0 10 20 30 40 50 60 70 q TIME (days) y Figure 8.2. Rete of Hydrogen Generatkn Based on NUS Code. i
= le "1 o.i 1 I 40 I l u I 'N Totat ' Pressure 20 k 1301 8 I 6 4 s/ [ /' ./ i 'H [ 02 2 N2 g 4 n.2 ,,/ c: b dg H /02 2 3,o 7 0.8 t i / r~~~ s 'J J' I O.6 I 0.4 o en 0.2 Mb 0.1 _ _J. I O 10 20 30 M 50 60 70 O'Yb WD . tis i f I l~sure 9.3. Prmu a Buildup in Contsimnent Follewing LOCA for ARC Sample Cam. B-6 8, 151<
m no-- + s <a-,1 n-aw .s. 1 1 t ..e t t (uiw/qi) 31YU Noll33fNI 2 = g o, I l j 7 l i i ] I i i i j l t .t 4.t 4 l 413 s l ii lilli a I. / l ~ 1 \\ ? I 8 i X a\\ l I -g i Y \\ i s 1 I e y I l( =0## l j l l\\ 1618 l / E3 k lllI I j i \\ ,na f O f* ~ l c 5' o,E k l o a o h k _ [ 2 !s 1 i l_ i\\ F 1 i ~5 i H 6 ! i ( c-s ! + i i I i h j l ) I f k 8' a i I l \\ / if ! I g{5 l \\ / gg;. i
- 2 l
\\ / i ! e m l '3s \\/ +- / I.q; / 1 + / a \\ i ..i'i _r i! v i e 4 _ / 1 m t E fl \\ r l-1 i / \\ a' t I i / \\ i,2 : i l [ l / \\ tiiill c / t l / l I l / e. O O O O j &7 l 155' J
- - _ ~. - - _. ~ lJ i i Shown also on 'igure S.4 is the changing rate in 1bs/ min at which 1301 3I would have to injected from time zero to insure that the required amount for inerting is present at all times. This rata drops with time and its highest value in 20 lbs/ min. Hence, in this example, icjection at a constant 20 lbs/ min l for 16.0 h'ours would deliver 19,200 pounds of 1301 and would insure protection f* at all timas.' Referring to Figure 8.3,'the Halon 1301 pressure rise and the total pressure for the sample case are plotted. The total pressure after 67 days is 35.2 psia, considering no steam. ij There is an alternative method of computing the Halon 1301 requirement ] as a function of time. The previous computation was based on the amount required d* to maintain the peak percentage of 1301 at all times for the instantaneous N /O 2 y ratio. It is also possible to compute the amount of 1301 required at all times 7 J. to maintain the H concentration below the lower fisemability limit of 4 percent. 2 This calculation is made by first determining the pressure of H and the total y pressure in the contai-nt,as.a function of time following a LOCA. These data are plotted in Figure 8.5 (for the ARC sample case). It is then a straight-j forward eniculation to determine the amount of 1301 that must be present at any time to' maintain the H concentration below 4 perce.nt. This required weight of y '[ 1301 is shown as the solid line in Figure 8.6. Note that at approximately 90 I' hours the required amount of 19,200 pounds will have been delivered. I* ' Shown also in Figure 8.6 is the injection rate at any time. This is O' obtained, as in Figure 8.4, simply by dividing t!.e instartaneous weight by the time. The maximum rate in this example is approximately 12 lb/ min. Since the actual containment inerting system would not have provisions for injecting 1301 at a variable rnte, a constant rate corresponding to the maximum of 12 lb/ min { would be chosen if this mode of maintaining protection were selected. The pres-sure rise curve for 13C1 injection to maintaa H e neentrati n below 4 percent 2 could be pincted on Figure 8.3. However, considering the extended time scale of that plot, the curve would be only slightly displaced from the line for 1301 that represents the peak percentage inerting procedure. The post-LOCA hydrogen concentration in the containannt is shown in Figure 8.7 (as percentage) with l Halen injection at a constant 12 lb/cin rate. Since this rate is greater than required to maintain 4 percent, the actual percent H; drops well below 4 percent by the time all the 1301 is injected. Shown also on Figure 8. 7 is the percent - 1 H that would develop if no Halon were injected. j 2 o 156< B-B ?
FI H H H H M M WI M M M M 9 "88 1 M M F8M 69 855 E'l55 I i l t I-5.0 1.0 0.8 / 4.5 7aS 4.0 0.6 I / 0 .r 9 2 C/l 0.4 3.5 3 mj / E e t 0.2 3.0 0 2.5 0 10 20 30 TIME (hours) f Figure 8.5. Partiet Pressure of 112 and Totel Pressure In Contelnment Following LOCA if no llalon 1301 is injected. i -J e
O I I ~ I 105 100 L / /
- 104
- ~'*-
/ 10 k i 7 l 1 v d ) ,,,,o I's 'T C /
- %9 g
/ t Wn 9 /, b ) w i ): 3: / z 1 i M o 9 E [ u f 5 . !d di $ 103 1.0 M i c i i-I + 1 ,t i [. J j*: / 11 j. ! {i I l a u l I l 102 o 1.0 10 100 1000 l l TIME (hours) i t I i ( e. 1. Figure 8.6. Halon 1301 Requirement For Containment inerting For ARC I* Sample Case, Based on Maintaining H2 Concentration Below 4.0%, i ~ B-10 i. 158 <- --m
!l ii h F. E O f se I 15 / [ / \\ No Halon 10 Ii b h l. 5 Halon Injection f M_ ~ L { 0 0 10 20 30 40 50 60 10 TIME (hours) i 1 Figure 8.7. Percent H2 n Containment Following LOCA with no i 1301 Injection and with injection at Constant 12 ib/ min. 8-11 i 153c i
=I -..,..,m -... ~. p~ n-l The foregoing discussion was intended to illustrate how the amount and injection rate of Halon 1301 is computed for a post-LOCA situation. In a real application, containment conditions would presumably be different than those-I selected.in the aussple case but the general C yrMch to computing the Halo'n re-quirement would be the sese. A concise picture ~ th. computation ae'thod is pro-1 vided by the logic diagran presented in Figure 8.8. This diagram essentially l represents a stemma7 of the computations previously mada. Two of the input con-J ditions were not considered in the example case, a ialy, the solubility rate and decomposition rate of 1301 in the emergency coch 4 water. When data on these { Processes become finalized they can be includad. e 8.3 EFFECT OF PRESDICE OF STEAM I The procedure fer cosputing the 1301 requirement has not included the presence of stesa in the contain==nt. Ultimately, the system design will have tc. taka steam pressure into consideration in. some way. At present there is no firu estimate of blevdown conditions, but it is still possible to describe the princi-Ples to.be followed in accounting for the presence of steam and to describe. qualitatively the effect. T [a, Pressure and temperature measurements will be made in a containment during a 1.0CA. From these measurementa, and, knowing the initial air contan e and [ the rate of H2 generation, the amount of stesa present can be computed. The best procedure for dealing with stesa pressure would then be to consider it as [ a diluent equivalent to nitrogen. From a combustion suppression standpoint this 4, would represent a conservative assumption. Steam would have the effect of de- { 1aying considerably the start of the injection of Halon, and would also affect 4. the final quantity required. { As an esseple, the plots shown in Figure 8.9 can be considered. These represent preliminary estimates of blowdown temperatures and pressures provided by Babcock and Wilcox. It is emphasized that these estimated conditions could b change considerably. At 15 hours the containment pressure is read as approxi-g estely 10 peig. This would consht of 0.55 psia of H,1.20 paia of 0, 3.72 pois l 2 2 [ N2 and thu remainder (19.5 psia) steam. Instead of a N /0 rati f 3.10, the 2 2 ratio would be taken as 19.4. The H e neentrati n would be 2.2 percent at this 2 i time. rather taan 10.1 percent which one computes with no steen present. For an if a i t 160< t-8-12 . - -. ~
y e
- -1
- ~l
.M M ._i . -. i s4 C U b3 t i 1 Containment J Leak Rate 'a f if t. I' Decomposition
- J lp!
Rate A = Maintain Peak Percentage For N2f02 and PTotal Solubility B = Maintain H2 < 4.0% Design Rate in H2O + Tonks 3, i e 1 P Compute Dry Well Compute 1301 1)
- Initia, Composition: H2 O2, Compute Halon 1301 Injection Rate Design b
Conditions -+ N2 as f (time).
+
Required as f (time) +
For Options
+
Injection P
2 A or B Nozzles
,f Compeste N /02 for Option: A and B M I y
Compute Time 112 = 4%
a' C C i!
H f
Engineering Estimates (To Right of Broken Line) l l
l l
IM 30 i
i i
i e
i 0
1 2
3 4
5 6
7 8
9 10 11 12 13 14 15 i
i HOURS 30 40 50 60 70 y
I DAYS' I
Figure 8.9. Preliminary Estimates by Babcock and Wilcox of Blowdown Conditions in Containment.
0 4
8
- n...
1 i = actual application, preliminary estimates of blowdown conditions would be made and Preliminary 1301 requirements computed. If a LOCA occurred, than actual 1301 ~ injection could be regulated by containment instrumentation which would monitor f existing conditions. The total' quantity of Halon required would be reduced sico if steam vers considered. For example Figure 8.3 shows a final N /02 **Ei f "PE# *i" 2 nately.0.85. If steam pressure of 1.7 pais (at 120'F) were present and taken te be equivalent to N, the final N /0 rati w uld become 1.3. According to 2 2 2 Figure 3.11, this would require at least 12 percent less Halan 1301 for inerting. j i 1 l, ma m 1tia< S-15 ~
_-._.y-..-...~.-.--+-
w.----------->
eem*~,: w& :.: = - _--~r: - v- :a i l i
9.0 CONCLUSION
S ll Thermodynamic calculations show that az;dosion and detonation presserres ( of post-14CA H N mixturne, if igniti n 6ccia red could exceed the contale 2 2 2 ment design. It is possigla to inert auch mixturas of any composition using Ealoa 1301. The required enount of Es.lon has been naamazed experimentally over a wide range of compositions and for various canditions cd temperaturn and pressure. Testing j was conducted on laboratory, intermediate, and large-scale. The results will te k used to specify the quantity of agent required for actual systems application. l Several potential problems that were postulated might arise in the app 11-j cation of a Halon system to a containment were investigated, ho of these - the i effect of 1301 on charcoal is the containment flitar system, and the radiolytic decomposition of gaseous Halas in the containment free volume - were shown to be g inconseq uen tial. However. another potential problem - the dissolution of 1301 im { emergottcy cooling water with stesequent decomposition as it passes through the reactor core - is still t:rmettled at this time. An experimental study within t t.he present program showed that the rate of sol billey of Halon into H O is quira 2 rapid. It therefore is necessary to investigate the rate at which 1301 decossposes in solution in the radiation field of the reactor core in order to datermine wtather appreciable Halon loss will occur over a 67-day period. This latter study is planned for the future. A sub-scale facility was assembled which simulated a 1301 syr tem in a containment application. Injection and vaportaatim tests of Halon were conducted I in this facility to identify any problems that mistt exist in discharging and j gasifying Isrgs quantities of the agent. On the besia of the resulte.cf the entire i program, with the exception of the unsettled H 0 decomposition question it is 1 concluded that a Halon inerting system is ideally suited for application to a a v i. j maritima reactor. A system analysis has shown tha:. to inert a 70,000 ft3 con-i 3 tainment of 33,000 f t ullage. 20,000 pounds (23 psia) of Halon 1301 would be required based on USNRC Reg.114 tory Guide 1. 7. e i \\ 164< I o-1 i i
- }}