ML17202V100

From kanterella
Jump to navigation Jump to search
SE Rejecting GE Methodology for Determining Rates of Generations of Oxygen by Radiolytic Decomposition (NEDO-22155).Rept Underpredicts Generation of Radiolytic Hydrogen for Boiling/Nonboiling Cases
ML17202V100
Person / Time
Site: Oyster Creek
Issue date: 07/06/1989
From:
Office of Nuclear Reactor Regulation
To:
Shared Package
ML17202V099 List:
References
NUDOCS 9011090439
Download: ML17202V100 (6)
v • d • e


Text

. -

UNITED STATES NUCLEAR REGULATORY COMMISSION WASHINGTON, D. C. 20555 SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION ENCLOSURE 2 GENERAL ELECTRIC COMPANY'S METHODOLOGY FOR DETERMINING RATES OF GENERATIONS OF OXYGEN BY RADIOLOYTIC DECOMPOSITION (NED0-22155)

In June 19S2 General Eleciric ~GE} issued the subject report containing a description of the methodology for determining rates of generation of oxygen by radiolytic decomposition of water in the inerted Mark I containments.

In this report, GE assumes that after an accident water in the containment will boil* for 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> only.

During this.time it will undergo radiolytic decomposition with oxygen generated at the rates corresponding to G(02 }=0.1.

Where G(02 ) is a number of molecules of oxygen generated by 100 ev*of radiant energy absorbed.

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

  • For radiolysis of water beyond 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />, when boiling ceases, G(02 )=0 was assumed and consequently there was no net generator or radiolytic oxygen; This last assumption was based on the.analytical results obtained by Knolls Atomic Power Laboratory (Reference 1} and by Argonne National Laboratory *

(Reference 2} in connection with the Three Mile Island accident.

The values of G(02 } ih:the GE report differ considerably from the value of G(02 } in

  • Regulatory Gui de 1. 7 which for botl:l boi 1 i ng and non-boi 1 i ng cases recommends G(02 }=0.25.

However, this value is not based on any specific mechanism of

.radiolysis. but is thosen to bound all possible cases and consequently it tends to overpredict the rates of generation of radiolytic oxygen.

In 1982 an extensive effort was undertaken by the North~ast Utilities and by the NRC in connection with the Millstone I licensing action<to determine a more realistic method for* calculating rates of 'radi.olyti.c oxygen generation.

In performing this task the staff was assisted by a consultant from BNL.

The results of

  • this effort: have indicated that.G(02 } is not a constant parameter but varies with the amount of hydrogen dissolved in water and with the concentrations of certain impurities, *most notable among them iodine.

Since concentrations of these substance*s may vary with time and may be different for different accidents,... *

.the true.value G(02 ) should.be expressed as a function of these variables.

In gener~l, an increase of concentration of hydrogen fo water results in a decrease of radiolysis due to promotion of recombination reactions.

On the other hand an increase of iodine concentration tends to promote radiolysis by qestroying free ra.dicals which are required for the recombination reactions to proceed.

The highest rate of oxygen generation is achieved when G(02 }=0.22;.

~*

.... which is the highest theoretical limit for gamma radiation.

This occurs when water is completely free of dissolved hydrogen, or when.the concentrations.of

.dissolved iodine are extremely high.

However, in most cases G(02 ) will be lower and at certain concentrations of hydrogen and iodine the rates of radiolytic

.dissociation and recombinations reactions may become equal -resulting in G(02 ).~0 and no net generation of radiolytic oxygen.

During the.boiling regime hydrogen.

will be stripped by vapor bubbles and it is e)(pected that G(02 ) will be higher

  • than ~n non-boiling_water.

Quantitative evaluation performed by the staff was based on the model developed by the BNL con~ultant (Reference 3) and on the experimental data from*ORNL (Reference 4).

For pure water (no iodine) it was determined experimentally that with no dissolved hydrogen and no boiling.G(02 )=0.08.

However, when under non-boiling conditions the concen~ration*of dissolved hydrogen reach_ed 2.5 cc/kg of water, corresponding to equilibrium hydrogen pressure of 0.16 atm.,

'G(02)**became zero and generation of radiolytic-*oxygen stops.

This finding

.contradicts the information in the GE report where G(02 )=0 was assumed for a 11 non-boiling cases.

  • For water containing dhsolved iodine no applicable experime.ntaf d~.ta were availabl~ and th~ staff calculated G(02 ) corresponding to the maximum credible

.iodine concentration in water using the BNL ~odel. Since all iodine in the cbntainment ~ater comes from failed fuel, an accid,nt had.to be postulated *

  • which would result in a release o'f this amount of.-iodine.' ln such an accident fuel was assu.med*to fail by oxidation of Zirconium cladding and hence, in
  • addition to re leased. iodine, additional hydroge~ was produced.

C"Crncentratj ons of both.the~e substances,had to be consfdered in calcu1at'in~* G(02 ).

The accident conside.red consisted of a LOCA in which. 5 percent of fuel cladding was oxidized.by rea~tion with steam produc_ing f~ilure of all fuel. rods and overheat_ing of ttie core, but without ini~iation of fuel me)ting.

This case represented maximum degradation of core allowed-~y 10 CFR 50.44(d)(l) and 10

. CFR 50. 46(b)(3). *, The analyses performed by Sandi a (Reference S), based on the

. experimental work on fuel rods from the.H. B. Robinson plant, have indicated

. that. for this type of accident 30 percent of tota 1 fuel iodine inventory was released*.

The.released iodine consisted of the* initial gap inventory and of the iodine diffused from the overheated fuel.

Assuming that all *the released iodine was dfssolved in water and u~ing:plant paramete~s corresponding to a~*

typic.al BWR with Mark I containment, the* iodine. concentration in water was*

determined to be 1.11 E-5 moles/liter and the partial pressure of hydrogen in the containment 0.12 atm.

This partiaJ pressure corresponds to an equi*librium concentration of 1.9 cc hydrogen/kg of water.

Inserting this value of iodine*

concentrati'on into the BNL mathematical model a relationship between G(02 _) and partial pressure of hydrogen in the containment was developed.

From this relationship it was determined that for a non-boiling case, whe~Lpartial pressure of hydrogen was 0.12 atm:, G(02 )=Q.19.

It also found that G(02 )

would not*reach zero*value until partial pressure of hydrogen in.the containment reaches i atm.

For boiling case, when hydrogen is stripped from the solution, G(02 ) would be slightly higher, somewhere between 0.19 and 0.22.

    • 'ir

y

  • )

~ -

These values differed considerably from those in the NED0-22155.report.

The main difference was *probably due to the GE result.s being applicable to pure water or to water containing only minimal amount of impurities.

Including the

  • effect of iodine, which would be released during.certain types of LOCA, could drastically change the results.

. CONCLUSIONS AND RECOMMENDATIONS

1.

The NED0-22155 report underpredicts generation of radiolytic hydrogen for both boiling and non-boiling cases.

This is due to the use of too low values for G(02 ). 'G(02 )=0.l for boning case was based on the measurements made in -an environment of zero or low iodine concentrations. G(02 )=0 for non-boiling case was derived from the data calculated by.the codes which did not consider effectS of dissolved iodine.* The results were* also in disagreement with the experimental data from ORNL. *

2.

Since G(02 ) is a fUnction of hydrogen and iodine concentrations -in the

  • .containment water, it may vary during an accident and is specific for each individual plant.
3.

The maximum valu~s of G(02 ), calculated ~ith the NRC r~di~lysis'model :for LOCA (5% metal-water reaction and* 30% iodine release) in a BWR with Mark I

.:containment, are G(02 )=0.19 for non-boiling and between 0.19 and 0.22 *.

for boiling cases.

They are considerably higher than the values presented

~-*in 'the General-Electric's NED0-22155 report.

4.. :.The vaJue of G(02 )=~25 in Regulatoi:y.Guide 1. 7 is ov*erly conser'~ative...

  • However_, it is not very much different-from the maximu.m value~ calc*u1ated
  • for.: a LOCA using the BNL model. *It 'is recommended, therefore that until
  • a better *understanding ~f post accident radiolytic decompo~ition of.wate~

.is developed, this va*lue should pe.used for predicting generation rates.

of radiolytic oxygen in. the* containment.. ***

REFERENCES-..

1.
  • J. C... Con.ine, D. J. Krommenhoek and o~.Emanual Logan, KAPL Evaluation of Radiolysis.Associated witn Three Mi-le Island Un.it 2 Incident,**dated May 1979.

.-2.

S.. Gordon*;.K. H. Schmil:ft.and J. R. Hanekamp, An Analysis of.the Hydrogen B~bble C9nce~n~ in the Three Mile I~l~nd Unit 2 R~actor Vessel~ Argonne National Laboratory.

3.

NRC. Memo and K. I. Parczewski to Victor Benaroya, dated **June 23, 1982~

4.

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

5.

Principal Contributor:

K.. Parczewski Dated: *July 6, 1989

~~~..Jl!B!

-.*.-**'lit***

~.

~****;*~,,

..~.

MEETING AGENDA FOR PISCUSSION OF PLANT SPECIFIC DESIGN OF.

ACTIVE COMBUSTIBLE GAS CONTROL SYSTEM ENCLOSURE 2 I.

INTRODUCTION BY NRC STAFF II. MEETING OBJECTIVE To provide sufficient. design details of the Nitrogen Injection Capability for each plant to determine 1f the prov1sions*are adequate to meet the 1ntent of 10 CFR 50.44.

III. AIR-CAD SYSTEM I

NRC staff w111 provide the basis for cc>ncluding. th~t the system should

  • not be used for :combustible gos controJ.

Represents an air source per GL s4..:09 guidan.ce Potential misuse of-system

. *.

  • Impact of d~inert1ng on c~uts~ ~f the accident r
  • IV. uci:N$ING. BASIS AND BWR EPGs Reconcile any confllcts betwe.en original' 11censing basis and Rev. 4_.

to the BWR EPGs~

Wherl wnl *,CAD be used (under.what system.co'1dit1ons)

Is the EOP con~istent with FSAR *as~umpt1ons

. V.

PLANT SPECfFJC DESCRIPTION OF INERTJNG, SYSTEM

. The objective* 1~ to.identify ill esseritial comp~nents, design conditions, instrumentation, and power supply for tht normal fot:rting system to..

. det~rmine under what post LOCA conditions.the sy.stem could be expected* to.

. fun ct ion as a combustible gas control system.

A.*

System/c~mponent description All compo~ents req~ired to ope~ate under p6~t LOCA ~onditions'should be described.

  • Desi9n specification (i.e. setsm1c, redundancy, quality group,

' etc.)' for* a parameter where the component has not been des tgned

. for, such as seismic, but 15 expected to survive, provide the

. basts.

Location

,~* ~*

~*1.,

. ~ *

. -*\\

- 2...

Design pressure and temperature Maximum containment conditfons for component oper~bf 11ty should.

also be noted

  • Design flow rate as a function of contafnment pressuf'.e B.

System Instrumentation lnstrumentat1on necessary to operate.the system post LOCA should be

  • identified. *.

Identify each sensor, nun~er and' location location of fnstrument readout (If outside control *room, determine acc~ssibi 11.ty post. LOCA)

System use of sensor output What"t~.sensor used for (autom~tit valve operat1on? or n9t). If operator :information only, indicate what type of action is

  • anticipated (flow. cha~ge, system shutoff, etc.J*

.,C.

Oper~tional require~ents post LOCA Idt:ntify tho.se actfons that will be* required to 1njt1ate sys~em oper~tion and those necessary 1*or monitoring operation. For *each*.

~ct ion, 'identify wh.ettier 1.t fs from the control room or at a *remote

.siti.*

lnstrumentat1on needed* for startup

  • instrumentation needed for cperatiol')

Power ~supply (offsf te and diesel?)

. Nitrogen capacit¥. (on~ite.and time t" get -addedsupply)

  • E.

ldentfffcatfon of devfatfons from GDC 41., 4?,* a~d* 43 F.

Modiffcatfons to eliminate devfations fr'om (E)

G.

Operatfonal, maintenance, and surveillance history - objectivt 1s to obtain *some.basis for detenn1.ning system *avaflabf11ty. '

-....

Retrieved from "https://kanterella.com/w/index.php?title=ML17202V100&oldid=5128304"