ML19290A258

Expresses Appreciation for Work Done Re Hydrogen Bubble Problem

ML19290A258
Person / Time
Site:	Crane
Issue date:	05/25/1979
From:	Harold Denton Office of Nuclear Reactor Regulation
To:	Slagg N ARMY, DEPT. OF, ARMAMENT RESEARCH, DEVELOPMENT &
Shared Package
ML19290A225	List: ML19290A224 ML19290A226 ML19290A227 ML19290A230 ML19290A231 ML19290A233 ML19290A234 ML19290A236 ML19290A238 ML19290A241 ML19290A244 ML19290A245 ML19290A246 ML19290A248 ML19290A252 ML19290A256 ML19290A258 ML19290A259 ML19290A260 ML19290A261
References
NUDOCS 7910240304
Download: ML19290A258 (4)

Text

,

Attachn:ent 2 WAY 2 51979 Dr. Herran Slagg, Supervisory Chemist Dynacies Brar s. Energetic Materials Division LCWSL U.S. Arn Arrament Research and Development Coarand Dover, New Jersey 07801

Dear Dr. Slagg:

We wish to express our appreciation for the asti;tance you, your staff and others provided to the Nuclear Regulatory C, mission staff regarding I

the hydrogen bubble problem at Three Mile Island-2 nuclear facility by calculating potential explosion pressures in tree upper region of the reactor vessel for the various postulated gas mixtures provided to you.

We realize that rany concerned persons worked evenings and weekends as well as regular hours to provide us with timely infomation.

It has since been determined that the available oxygen for detonation or deflagration of the hydrogen was considerably less than assumed at the time, however, this does not lessen our appreciation of your efforts.

Please convey our thanks to your associates and to Dr. John Kurylo and Pmfessor A. X. Oppenheim of the Lawrence Berkeley Laboratory for their repott on this r.atter.

Sincerely, c.tej nyW t

.. n....

Harbid R. Denton, Director Office of Nuclear Reactor Regulation N

2226 206 7910240f d

BROOKHAVEN NATIONAL LABORATORY MEMORANDUM OATE: April 4,1979 To:

W.Y. Kato FROM:

John R. Weeks

/

SUBJECT:

Discussions Relative to the Three Mile Island Incident cj,\\

1.

Hydrogen in Containment h

Walter Butler of NRC asked me to estimate being pcssible the build-up of

1. gg[ His rough guess was that e

hydrogen in the containment by radiolysis of water in 2 high y field.

I in turn discussed it vitli Dr. Harold Schwar: of the BNL Chemistry Department.

M the hydrogen may build-up to several percent which should be approaching the ignition point. The higher the te=perature (above 100 C), however, the greater would be the recombination rate and the less the build-up of hydrogen.

2.

Discussions Concerning the Hydrogen Bubble in the Reactor Vessel grA Warren Ha:elton asked me what infor=ation I had on the thermodynamics and e hkinetics of the reaction of hydrogen at a high te=perature and pressure inside j[d I the reactor vessel on the possible decarburization of and methane formation in J

the vessel material.

I discussed this subject with David Gurinsky and J. Chow Y

p of BNT., M. Gensamer, yrofessor Emeritus at Columbia and A. Ciuff reda of Exxon Research.

The stainless steel cladding on the inner surface of the vessel would be a partial barrier to hydrogen provided it were ntact.

There is enough of a chance of a flaw in this cladding, however, that no credit should be taken for it in estimating the perfor=ance of the reactor vessel material. The reactor vessel is =ade of a pressure vessel steel (ASTM A-533-B) which contains 1% Mn, 0.5% Ni and approx 1=ately 0.5% Mo.

The oil industry is continuously concerned about hydrogen induced decarburization of steels in their refinery equipment.

They have prepared a graph stating the safe temperature and pressure for steels (Nelson Diagram) in the American Petroleum Institute report API-941, which was most recently modified in 1977.

A steel of the composition used in the Three Mile Island vessel should be safe from decarburization by 1000 psi of hydroge. at temperatures 0

up to 700 F for indefinite use.

Exceeding this te=perature or pressure for short periods would not cause serious damage as there is a definite incubation time, of a =atter of several days, bef ore problems begin to develop.

Mo appears to be even

= ore eff ective than Cr in retarding this decarburi:ation although the reasons are not clear.

The same steel without the Mo would only be saf e up to 500 F at 1000 psi of hydrogen.

I think the upper part of the reactor vessel should be carefully checked for any possible da= age from decarburization prior to its return to service.

A ccpy of the curve showing this relationship as revised in 1977 is appended to this ca=orandum.

2226 207

e4 W.Y. Kato April 4, 1979 Hazelton also asked whether radiolysis of the water within the vessel could add oxygen to the hydrogen gas bubble.

In my opinion, it should not.

Radiolysis of water p-eeds by a complex chain reaction and can be prevented even by a s=all overpressur, of hydrogen in an operating FWR. The high hydrogen pressures over the coolan at Three Mile Island should totally prevent oxygen formation.

In fact, Harold Schwarz stated it may be feasible to recove the hydrogen by simply adding oxygen slowly to the coolant; this could, ad=1ttedly, be risky.

I think we should be very careful not to use chemicals such as sulfate or sulfur bearing compounds to react with the hydrogen since these can be reduced by the excess hydrogen to sulfides which are very harmful to a number of the materials in the system, especially the Inconel steam generator tubes.

It might omplicate the return of the unit to service.

I recommended that a nitrate (suc' as potassium nitrate) be used if one wishes to go by this route.

However, I think the best means of hydrogen removal would be through venting it from the primary coolant into the containment where it can be recombined with oxygen.

3.

Some Crude Calculations of the Amount of Zircaloy that Participated in a Zr-y0 Reaction During the Incident

.f l I estimate that as much as 3200 lbs. of Zr may have reacted with water to 0,'.$pp produce the hydrogen bubble, assuming it occuppied 750 cu. f t. at 500 F and g

1000 psi, as stated by Hazelton.

This suggests that over 10% of the Zircaloy

,Ij yl cladding in the core was converted to oxide by reaction with the water. Whether or not the remaining Zircaloy could act to remove hydrogen from the water by

[ p hydride formation is not clear.

However, the hydrogen overpressure during normal PWR operation does not cause significant hydriding of the fuel cladding so that hydrogen removal from the bubble by this mechanism see=s unlikely. This hydrogen (10-50cc STP/kg H O) amounts to a maxi =um of 3.24 lb. in the pri=ary 2

coolant (329,200 kg) so clearly, the majority of the hydrogen bubble came from sc=e other source such as Zr-E 0 reactions, if the bubble was as large as 2

described by Hazelton on 3/31/79.

JRW:ob Distribution FJL J. Chow D. Gurinsky H. Kouts NRC V. Noonan W. Hazelton'

?. Almeter 2226 208

i STEELS FOR HYOROGEN SERVICE AT ELEVATED TEMPERATURES AND PRESSUAES 5

HYOROGEN PARTIAL PRESSURE. MPs abe 14' 1.38 2.07 2.76 3.45 4.14 4 83 5.52 6.21 6 90 1000 C'L

-a 2

~

12 11

'3 l 2

Es Es N'

t*'.,

5l E3 ang-

~

• 8 y15 i

m 14 8 13 1""

,, 1

\\

14

\\E= l En 15 C

14 s 34=

.g 9 T6* ~8 U

m x

s.

r=5 N-400 g s.

I C E3 d

e9 a

8 8

\\

700 0

g

^

1-l N 1

N 6g 0.5 Mo STEEL k

\\

h i

i i

I s

k

\\

0.25 Ms STEEL

\\

9C c7 h

l l

)

08

'9 eg 300 l

89 e9 8,3

,j l

l 0.1 Mo STEEL (1400)*

CARBON STEEL' g

l I

I I

~-

500 100 200 300 400 500 600 700 800 900 1000 HYCROGEN PARTIAL PRESSURE. Lb/h* abs

EGENO References Analysis Mo Equivalent CR Me Y NN O.01-c.i n-c21 cat-0.41-0 51-

$7 0

a.io amo cao c4o oso oso I. Shell Ost Co.*

0.50 0.50

2. Weld Ceposits. D.J. Bergman

• 0.79 0.J9 0.39 same o

0 0

1:2 8 t:2 g

3. Weld Deposits. D.J. Bergman

• 0.30 0.15 0.35

.nonocee arTacx e a

I E.

E &

M

4. Weld Deposits. D.J. Bergman

• 0.50 0.25 0.37

3. Continental Oil Co.*

0.25 0.25

6. Standard Oil Co. of California

• 0.27 0.27 Standarc Oil Co. of California

• 0.05 0.06 0.08

3. A.O. Smnh Corp.*

0.13-0.18 0.11

9. Sheil Destiopment Co.

Draing No. VT 639-2

10. AMOCO Oil Co.*

0.04 0.0t NOTES:

11. Ct.W \\ta nuel. Corrosson.

1. \\1o has four times the resicance of Cr to H: attack.

I? !91. pp.1034 Sept. 1961 0.27 0.15 0.22

. \\1o is coun alent to v. Ti. or Nb up ro o.1 percent.

. The Star. card Oil Co. of Ohio

• 0.11 0.43 0.50

3. Si..Ni. Cu P. and S co not increase reustance.

3. Enon Cors.*

a. :- Ca af Cahfornsa*

g.]

I g

.f. A e o cii Co.*

r hI

- J. JC' r' g 9 b card Oil Co. of Ca:ifornia*

1. gg C..? O, Carp.*

j 1977 -

ce,e-- u.. cation to Subee-me:ee on Carrouon

.o.

igwre 2-Ocerating Limits for Steels in Hycrogen Service Shcwdg E"c;t Of Trace Alicying Elements 2226 209