ML19345E393

Forwards Addl Info in Support of Proposed Change 8 to License DPR-6,per 660124 Request.Refueling Shutdown of Facility Scheduled for 660415

ML19345E393
Person / Time
Site:	Big Rock Point File:Consumers Energy icon.png
Issue date:	03/17/1966
From:	Haueter R CONSUMERS ENERGY CO. (FORMERLY CONSUMERS POWER CO.)
To:	Boyd R, Doan R US ATOMIC ENERGY COMMISSION (AEC)
References
NUDOCS 8101160249
Download: ML19345E393 (18)

Text

.

_m (g,%

Consumers k 7"3 Power

/

Company orne.n. omu. : s i 2 wo u u.e,,9.n 4.cnue..mo.on. u.oi,en v.2v t. Area C..M

• 11 18 0 - OMO March 17, 1966 Dr. R. L. Doan, Director Re: Docket 50-155 Division of Reactor Licencing United States Atomic Energy Commission Washington, D. C.

20545 g

opy

Dear Dr. Doan:

Att: Mr. Roger S. Boyd Trancmitted herewith are three (3) executed and nineteen (19) conformed copies of additional information in support of our Pro-posed Change No. 8 dated December 23, 1965 This additional information was requested by you in your letter of January 2h,1966.

The refueling shutdown of Big Rock Point is currently scheduled for about April 15, 1966.

Yours very truly, us/

RLH/dmb Robert L. Haueter Ene Assistant Electric Production Superintendent - Nuclear

\\ pt

.~ %

,} \\% ^ q j.&,.

, (QLQ/

M

,r a

V

.. %\\

g? oorn m'p' ifs,x%

cn/g-

/

49 am M

t w 21 19% ' $

N J b i.M h.

~. /4

/f f;y..,y,gA mamu at stena e

W

< e,..

q cecm can y

,?

a

'\\

N e

m Q

7(Oll&D A

~ '

kcrF"lMOE0 L 4.!

7., 3,..... g

,. w g. ~

~

2 I.. Thermal Hydraulic (Contd)

'The present-operating limits of the Technical Specifications will of

- course apply to any mixture of initial core fuel', R&D fuel and reload 1

fuel that may be used.-

The evaluation of the reload fuel (with respect to nuclear and thermal-hydraulic considerations as summarized in " Proposed Change No. 8") was performed by examining "most severe" operation conditions and then making a determination that the limits of the Technical Speci-

]

fications could be met within the range of possible core 1cadings and that it was quite likely the various peaking factors would permit achieving full power.

In direct response to your question, we have developed a core typical of the next loading and a core typical of the equilibrium core several years "down-the-road." In order to supply the detailed numbers requested, we have performed (on these two typical cores) the i

same type of calculations which we will perform later to provide as with f

the necessary information on the specific core loadings involved.

After selacting and defining an expected core pattern, physics studies provided the core power listribution necessary to carry out the thermal-hydraulic evaluatic.is. The thermal hydraulics were basedupon standard General Electrii. Company APED computer codes.

These codes consider the core flow distribution, power distribution and steam-water pressure drop effects. The methods are well developed an-4 alytically and have been confirmed by extensive reactor experience.

The following data are typical for the two 84-fuel bundle cores under consideration. The appropriate power distributions were evaluated for each of these cores.

All Zr-Clad Next Core Equilibrium Core Total Core Flow Cross Section, Ft lb.83 13 96 i

Total Recirculation Flow, 10 Lb/Nr 12 3 12 3 HotpannelCoolantFlow-Overpower, 4

10 Lb/Nr 0.12h 0.123

' Core Pressure Drop, Rated, Psi 55 5.h i=

t

~

~

.-.,e s.e,,

,,,--,r.

,nn e---,...r,w--m-,a

3 l

I.

Thermal Hydraulic (Contd)

AllZh-Clad Next Core Equilibrium Core Average Heat Flux, Btu /Hr-Ft Rated Power 113,000 113,000 l

overpower (1.22) 138,000 138,000 Peak Hesc Flux, Btu /Hr-Ft 307,000 2h8,000 l

Rated Power-37h,ooo 302,000 overpower (1.22) l Peak Linear Heat Generation, Kw/Ft i

overpower 12 9 lo.h Minimum Critical Hea*,1Jux Ratio j

Rated Pover 2 38 2 94-overpower 1 95 2 57 Maximum Fuel Center Temperature, F Rated Power 2,820 2,34o overpower 3,h80 2,770 Power Peaking Factors Radial 1.27 1.26 Axial 1 39 1.25 Local 1 53 1 39 Burnup at Start of Cycle, Mwd /T Peak Bundle 7,hoo 13,600 Core Average h,500 5,70o It may be noted from this table that every factor for either core is well within the present operating criteria by a substantial margin. All intermediate core configura-tions will be similarly chosen to meet the required operat-ing limits.

Question 2: Discuss the method for orificing the fuel. Will the orifices be changed as the fuel is moved within the core?

Answer: The design criteria for initial fuel, development fuel and reload fuel have resulted in essentially equal nuclear characteristics, except for exposure lifetime. Similarly, the same thermal-hydraulic i

limit criteria have been applied to the various fuel types. These similar nuclear characteristics have permitted the orificing of the fuel by lo-cation in the core only, irrespective of the fuel type. Thus, the core is orificed in two zones only with smaller orifices in the periphery and a standard orifice throughout the rest of the core.

I

h l

I.

Thermal Hydraulic (Contd)

The current (or first batch) reload fuel is alco very similar to the fuel presently in the reactor in both nuclear and th'ermal-hydraulic characteristics. Thus, the reload fuel does not require a change in this i

orifice pattern to obtain the best calculated thermal-hydraulic advantage within the stated operating limits. Should core calculations for subsequent

-cores indicate improvement by altering the orifice size or orifice pattern, it will be done in a similar two-zone arrangement and in full compliance with the operating limits imposed by the Technical Specifications.

Question 3: The calculated temperature coefficient at 77 F is an order of magnitude greater than that listed for the original fuel. Give the tem-perature coefficient at power and reevaluate the " cold water" accident.

Answer: The moderator temperature coefficient at 77 F and the void coefficient at 58 F were both in error in the December 23, 1965 submittal (Proposed Change No. 8). We should have caught this error since the nuclear characteristics of the reload fuel are very similar to those of earlier fuels.

Thus, the moderator temperature coefficient would have to be positive and small at 77 F.

The correct values are as follows:

Moderator Temperature Coefficient (bk

/k per F) ff eff 77 F

-6 Start of Cycle

+3 2 x 10

-5 Fad of Cycle

+5 5 x 10 VoidCoefficient(dk

/k per Unit Void Within the Channel) df df 0

68 F 572 F

-0.04

-0.09 With the above corrections, there is no significant difference in the temperature coefficient from previous fuels and the previous analysis of the cold water accident still applies.

II.

Nuclear Question 1: What is the calculated peak rod worth with the new fuel configuration? What is k without rods (hot clean, cold clean, hot df equilibrium) and k with the strongest rod withdrawn from the core (cold df clean)?

p 5

i II.. Nuclear (Contd)

Ancver: Two representative cases were calculated to de.termine peak rod' worths; namely, an estimated core configuration for the next refueling and a typical equilibrium all Zircaloy-2 core. The maximum rod worths were calculated to be less than 10 percent greater than the worths used in the various accidents analyzed in the Final Hazards Summary Report -

(FHSR)(0.0k2bk). Because of the tenuous na+ure of such analyses in general and also considering the very small p..bability of the abnor=al control rod withdrawal sequence necessary to n 41op such high rod worths ever becoming a reality (several procedural ec

)le must be violated deliberately), we do not consider this to be a nificant change. Addi-tional considerations, such as the fact that the

'esent analysis was done based en the first reload fuel batch, current on site, which is more reactive than vill be necessary for future rea sds, serve to re-i inforce this conclusion.

Zie following table lists results of the alculations for the tuo representative cases:

4. ' ' Zr-Clad Equillurium Core Next Core (Based on First Reload l'ypical Core Design)

MaximumCalculatedRodWorth,bk O.043 0.046 k,ff HotEquilibrium,RodsOut,bk 1.05 1.05 k

HotClean,RodsOut,bk 1.11 1.11 g

ColdClean,RodsOut,bk 1.12 1.12 keff k,ff Cold Clean, Skongest Rod M,bk <0 997

< 0 997 With the strongest rod out, the cold clean k,ff is operation-ally verified to satisfy this criteria to be less than 0 997 i.

Question 2: What testing vill be performed to ensure uniform nuclear properties in the reload fuel? What precautions are taken against loading a high enrichment pin in an outer position in a fuel bundle? What are the consequences of this misplacement?

Answer: The reload fuel, as part of the normal quality control tests, is verified as to uniformity of nuclesr properties by critical tests prior to shipment. Stringent procedural controlc are imposed during fabrication to route one tyg fuel rod to the assembly area for each loading step. The r

1 c

r.,

\\

6 II. Nuclear (Contd) end plugs on the fuel rods are shaped differently for each' of the two en-richments to assure easy identification 'during aceembly.

A~ separate quality control inspection confirms the loading pattern by means of the easily visible end plug markings.

The probability of an assembly loading error in which a high enrichment fuel rod occupies a peak power peripheral location in the bundle

-is extremely small. Its ultimate application in the reactor further diminishes the probability' of significant error. First, the above procedural controls and quality controls must both fail; second, the error in loading must specif-ically place the high enrichment rod in a peak power location in the bundle; and third, the bundle thus improperly assembled must be located in the peak bundle power location in a core of 8h bundles.

If, after all these precautions, a high enrichment rod were i

to occupy a peak peripheral region in a peak bundle, it is calculated that I

it would experience a h3 percent increase in heat flux. At licensed rated conditions, the MCHFR would drop 43 percent for the misplaced rod, which for the calculated 1.22 overpower condition would still give a value of 1.05 for the MCHFR and burnout of the rod would not be expected. At conditions of the maximum licensed overpower heat flux of 530,000 Btu /hr-ft (which is near tite calculated threshold

• of fuel center melting)', this hot rod, being h3 per-i 4

cent higher, would probably undergo central melting.- This center melting i

could possibly result in fuel and clad swelling with potential cladding rupture. It should be pointed out, however, that with the Bh-bundle core, the present overpower peak heat flux of approximately 400,000 Btu /hr-ft 2

is some 40 percent less than this licensed 530,000 Btu /hr-ft. Hence, with I

any practical 8h-bundle core configuration presently conceivable, including that \\'ith reload fuel, it is improbable that central melting would actually I

occur in the misplaced rod even at overpower.

In both the MCHFR and central melting situations, the potential

. consequence of the rod misplacement is clad rupture. The effects of clad rupture are already a matter of considerable experience in the case of stain-

'less steel clad fuel and are fully within the safe procedural handling in-corporated in regular plant operational procedures. Although inconvenience i

may result, no damage to core and internals is likely should the improbable l

e w.

.e.

,.,c,

7 II. Nuclear (Contd) fuel rupture actually occur due to rod misplacement. It 4s most likely that the failure vould occur in a peripheral rod and could b'e located by visual examination.

III. Accident Analysis Question 1: If the control rod or fuel bundle worths have changed significantly in the intermediate or full reload cores, what are the re-sultant effects on the hazards considerations for the fuel drop and rod ejection accident situations?

Answer: Various analyses indicate that the control rod or fuel bundle worths have not changed significantly over those cases analyzed in the FHSR.

The equilibrium all Zr-clad core was analyzed for a bundle wortn of 0.01h bkascomparedtoaworthof0.012bkusedintheFHSR.

For the same initial conditions, the peak fuel temperature was 3300 F, whereas in the FHSR accident analysis, it was 2500 F.

Both of these temperatures are below normal operating temperatures and hence are well below the level of-significance. Likewise, for the rod drop-out accident, the difference be-tweentheequilibriumallZr-cladcorerodworthof0.046bkandtheFHSR worthof0.042bkcannotbeconsideredtobeasignificantchange.

Question 2: Give the justification for terminating the metal-water reaction at 3300 F rather than assuming the ZrO mintains the clad in-2 tegrity to the melting point of the oxide in the coolant loss accident.

What additional reaction would take place in the latter case?

Answer: All safety evaluations of reactor performance which are con-ducted by General Electric Company (GE) on the general subject of core de-i struction and associated metal-water reactions involve the following as an important boundary condition:

Zirconium metal upon melting in the core region will fall from that Ngion unrestrained by any associated zirconium oxide formations.

The basis for such a boundary condition is presented in the follovira discussion on this subject by qualified technical experts in this field employed by GE:

"The termination temperature for the metal-water reaction has been taken as the melting point of the cladding since 4

,---w e

e e.-

m- - - -r

, - - - + -

--w-

.=_

I 8

4-f III.. Accident Analysis (Contd) at least 1957 when Lustman of Westinghouse (1} used this concept in the analysis of the loss-of-coolant

• accident.

The same basis was employed by Owens, et al, (

at GE-AFED in 1959, and essentially by all workers in this field until 1965 Experimentally, there appears to be no question that when zirconium, oxidized by steam, is heated above the fu-sion temperature, the resulting molten metal drips away Lemon and his associates at Battelle(3) in 1957, rapidly.

Furman and McManus at Vallecitos in 1960 and workers at GE-APED have all succeeded in melting zirconium metal in a steam atmosphere with no evidence of hold-up of the molten metal by a crucible of the ZrO xidation product of the 2

reaction. Experimental work currently being undertaken at General Electric Co. involving prototype testing of fuel rods under accident conditions also confirms that the metsl i

falls from the heated area upon melting. The only known ar-

^

gument for this crucible effect was a recent observation by Baker and Ivins(5)(6) at ANL. These workers heated Zircaloy clad rod in 15 psig steam to a maximum temperature of 2732 F l'

for about th:-ee hours and, from the amount of hydrogen evolved, concluded that about 74% of the cladding had reacted. Of this j

experiment, which was carried out at a temperature some 630 F below the melting point of the metal (3366 F), they stated:

J:

The oxidized cladding kept its original shape, and except r

for some cracks, retained considerable strength. It is likely, therefore, that the partially oxidized cladding will not collapse at the melting point of zirconium.

Interactions between Zircaloy, i.e. zirconia, and the fuel could lower the melting point of the fuel to below 5072,F; however, this has not been established experimentally with I

any certainty. Rapid collapse of fuel rods and significant downward flow of material in the core will probably re-quire temperatures close to the melting point of Zr02 (4718 F) or UO2 (5072 F).

]

I 4

1 ca-Wr--

~

e'r sy--

p

2. +-

av m

-y s;-rw4

-yr 5

~

9-III. Accident Analysis (Contd)

"It will.be noted that the basis for this stat ~ement is not experimental' observation at and above the melting point, but is a speculative extrapolation from lower temperature data.

Further, an examination of the photograph (6) of the Zircaloy-2 i

fuel element shows that, in fact, cracking and spalling of the oxidized layer was quite extensive. This cracking might be r

sufficient to preclude the possibility of the crucible effect at the melting point of the metal.

' "Zr0' is notoriously unstable at high temperatures; deliberate 2

attempts to produce a crucible of this pure ceramic have been consistently unsuccessful. This is due to the fact that at about 2201 F the oxide undergoes a solid-state transition ~

Zr02 (mon clinic)

> Zr02 (tetragonal) with evolution of heat (7,8,9,10,11), lh20 calories per mole, and' a large change in volume, about 9%, Ryshk2 witch (12, p. 3F3) states:

.With the trancformation from the monoclinic to the tetragonal modification, zirconia undergoes a con-siderable volume contraction, and vice versa, amounting i

to 9% by volume. This volume instability of zirconia, particularly apparent at the ' ransformation temperature, t

makes it understandable that firing pure zirconium oxide cannot produce coherent strong pieces of sintered zirconia.

Man: attempts to manufacture pure sintered zirconis failed I

invariably and completely, indeed, Only zirconia with a e

considerable amount of contaminations or additives could be fired to comparative 1, strong and stable ceramic pieces......

"The stabilized zirconias referred to above generally contain significant amounts of Cao, MgG or Y 0.

hese oxides in solu-23 tion result in a decrease in the equilibrirn transformation 4

temperatures to values so low that reaction kinetics are too slow for sir,nificant transformation to occur; and the meta-stable cubic form is retained to low temperatures.

.?

10 w

III. Accident'Anarysis(Contd)

"In view of these manifest difficulties in pre aring h pure Zr0 rucible intentionally, it seems doubtful'that a ctable 2

crack-free shell of thisi oxide could be formed spontaneously, under accident conditions.

"It should also be noted that in the Zr-Zr0 system, the high 2

melting point of 4716 F can be achieved caly with pure Zr0 '

2

.The presence of any Zr 6etal in solution (it is extremely soluble in the oxide, (I2, p. 387), up to 5% by weight. at room temperature) lowers the melting point, as can be seen from the phase diagram g(ven by Hansen(13)

The eutectic for the Zr-ZrO pair occure at about 35 mole percent Zr0, and a 2

2 temperature of 3452 F.

Retaining molten Zr in an oxide cru-cible above this temperature would not seem possible.

"There is experimental evidence,-consistent with this con-clusion, that temperatures approaching the melting point of Zr0 will be difficult, if not impossible, to achieve. Some 2

limited experiments on actual fuel elements have recently been carried out at Oak Ridge

........by melting Zircaloy-clad 1,0 fuel specimen in 2

air in induction heating.....

"For this ternary mixture of Zr-ZrO -J0, a melting point of 2

2 3902 F was observed. The authors atate (14, p. 38) that the formation of thdse low-melting reactor component mixtures seems to imply that fluidization of portions of the 'eactor fuel could occur, under some circum-stances, at much lower temperatures than have generally j

been assumed.....

"This conclusion would appear to be supported by both observa-tion and theory, and is consistent with the method of calcula-tion which has previously been used, rather than the higher cutoff temperatures suggested by Argonne."

The basic experimental work on this subject now being conducted at GE has offered considerable confirmtstion to this premise. In all cases in which fuel rods were heated up thrcugh the melting of the cladding, even

r 11 i

III. Accident Analysis (Contd) with zircorf.um oxide on the cladding surface, no crucible"effect was ob-served. The melted metal quickly ran down from the test sections and was

. quenched in water reservoirs below. Tne quenched residue revealed metal percentages conristent with the percentage used in the submittal.

It is obvious that the question of maximum in-core temperature of the reacting zirconium is very critical in the conduction of the usual metal-water reaction - 100% core melt analyses. For instance, the use of zirconium oxide melt temperature rather than the apparently correct zirconium melt temperature would cause the quoted percent reaction for the entire zir-conium inventory to increase from approximately 27% to approximately 60%.

1 Therefore, it wat of grave importance to not only investigate all of the prior basic work, but to also conduct actual prototype testing in which fuel elements vere melted under simulating accident conditions to confirm our

. accident analyses basis.

Question 3: our calculations indicate that if the hydrogen released in the metal-water reaction were to pocket and recombine, thus adding energy directly t, the containment atmosphere, a recombination of an amount of hydro-gen less than half that evolved from the calculated 2T% reaction would be I

sufficient to bring the containment to the design pressure. Similarly, if the hydrogen burned as it entered the containment, the energy would be added directly to the containment atmosphere, but some heat would be transferred to containment heat sinks since the combustion is assumed to take place over about 15 minutes. In this latter case, our calculations indicate that a full recombination of the hydrogen still could not be tolerated. Please provide your analysis of there two cases of hydrogen recombination.

Answer: Assuming that the hydrogen burns as it enters the containment, our analysis indicaten that recombination of 75% of the hydrogen evolved in the 27% reaction would be necessary to bring the containment to the design pres-sure of 42 psia. This must occur during the first 900 seconds since the building spray comes on at 900 seconds. Metal-water reaction calculations show that only 67% of the hydrogen actually will have been made available through. reaction of water with zirconium (ne18% metal-water reaction, Figure 2 of submittal). A containment pressure of 41 psia after 900 seconds would result, therefore, if no credit is taken for heat transferred to containment F

i

... ~

12 t

III. Accident Analysis (Contd) heat sinks. After 900 seconds, the incremental energy adilition caused by the hydrogen burning vill merely have a sloving effect on tile rste. of:

pressure decresse resulting from the containmer.t spray (higure 3 of.sub-mittal).

Transfer of energy from the atmosphere to containment heat sinks will, of course, reduce the above pressures and pressure rates cor-respondingly. Considering only the containment shell as a sink by con-vection during the first 900 seconds, and assuming a total hydrogen release -

and burn, approximately 12% of this energy is calculated to be removed from the containment atmosphere. For the 100% burn in 900 seconds, of course, this is not sufficient relief and the pressure would reach approximately 43 5 psia. But again, this is an unrealistically conservative assumption since a maximum of 67% of the hydrogen reaction can be completed within

~

this period.

Since the temperatures vill be lower in the more realistic 67% reaction case, less than this convective heat transfer to the contain-ment shell would occur. However, it is likely that other sinks such as the fuel pool and internal structures, combined with the shell, vill absorb undefined amounts of energy approaching or exceeding this calculated level.

Assuming then, a comparable 12% transfer of energy from the containment atmosphere in the case of the 67% burn during 900 seconds, the pressure vill be belov 40 psia.

The above discussion assumes the reality of recombination of all the hydrogen as it is generated from the reaction. It is our conten-tion that this is very improbable because of the containment shape, hydro-gen diffusivity and action of the containment spray as discussed in connection with Question III-4.

If it is arbitrarily assumed that the hot hydrogen pockets in such a manner as to establish a combustible mixture and it is further assumed that an ignition source is present in the poe t, recombination of the hydro-gen with oxygen vill take place. A con::ervative evaluation of the resulting pressure can be made using the following calculational method: All the hydrogen is assumed to recombine thereby maximizing the energy released.

No energy is lost from the gas mixture during recombination. No vn+er evap-orates into the= mixture during the recombination cf the gases. Finally, it

=

13 III. Accident Analysis (Contd) is assumed that the recombination takes place when the ccntainment is at the peak pressure of 35.6 psia and peak temp <ature of '223 F as shown in f

. Figure 3 of the submittal at 900 seconds after the accident. This method maximizes the resulting pressure. Under these conditions, the pressure rises to 45 psia and temperature to 440 F.

Using the same tethods and conditions ar above, but only recombining 3/h of hydrogen, will result in a pressure equal to design pressure (h2 psia at 356 F). A similar re-sult is obtained if all the hydrogen recombines but only 3/4 of energy released upon recombination is retained in the gases. In other vords, if 1/h of the energy released is transferred to sinks in the containment, de-sign pressure will not be exceeded. The fuel stcrage pool, the structural and radiation shield concrete and the containment sphere could act as such heat sinks. For instance, the containment shell has approximately 3 ;000 ft of surface exppsed to the contained gases. The shell is 0 7 inch thick and the portion exposed to the contained gases weighs approximately 1 5 x 10 //m.

A temperature rise in this mass of 13 F will absorb 1/4 of the energy re-leased by the recombination. It is reasonable to assume that the shell is not above 223 F at the time the recombination begins since this is the j

temperature of the gases at this time. At the end of the recombination proc-ess, if 1/4 of the energy is lost, the gases are at 356 F.

If the recom-bination is assumed to take place over the first 15 minutes after the ac-cident, the shell would be exposed to gases which rise 133 F in 15 minutes.

Under these conditions, a natural convection heat transfer coefficient between the gases and the shell will not exceed 1 Btu /hr-ft F and about1/8oftheenergywouldbetransferredtotheshell. The conclusion is that probably not enough energy could be transferred during the postulated f'

15-minute continuous burn to stay within containment design pressure if only shell heat absorption is connidered.

It should be noted that neither the FHSR nor the above analyses take any credit for the large heat-a.bsorbing capacity of the spent fuel pool.

This was done due to the complexities of the calculations involved. However, this is indicative of the very conservative nature of such calculations.

For instance, an 80 F, rise in the temperature of the fuel pool water will 7

absorb 7 x 10 Btu, the equivalent of all the energy contained in the pri-mary system fluid.

14 III. Accident Analysis (Contd)

Based upon such considerations as this, we conclude that containment design pressure vould not be approached under either of the postulated cases.

Question 4: In light of the above question, discuos the desirability of eliminating the present 15-minute delay in the operation of the post-incident containment spray. If initiated during the first minutes of the accident, would one set of sprays (400 gpm) be adequate to remove hydrogen combustion heat if the hydrogen burned as it entered the containment?

Could the containment spray be expected to prevent hydrogen pocketing?

Answer: One set of sprays (400 gpm) is sufficient to absorb the re-combination energy. The mechanism would be one of evaporating water from the spray into the hot gases. If the gas mixture is kept saturated with water vapor as the recombination occurs, the resulting pressure is 39 rsia.

The evaporation of an amount of water equal to 70 seconds of spray (at 400 gpm) vill limit the pressure to 42 psia. Therefore, it is reasonable to expect the spray to be able to cool the containment gases.

It is difficult to discuss the effect of a containment spray on possible pockets of hydrogen. It is our contention that significant hydrogen pocketing is unnatural because of the shape of the containment and the high diffusivity of hydrogen. The action of a containment spray in promoting turbulent mixing and natural convection currents is to be expected.

However, this action does not seem to be any stronger an argument against hydrogen pocketing than the containment shape or the high diffusivity of hydrogen.

Figure 2 of the December 23,1965 (Proposed Change No. 8) submittal shows that the metal-vater reaction takes place in a 30-minute period. The 15-minute hydrogen release period used in the calculation of peak containment pressure was taken as a convenient and conservative period.

Convenient because it releases the hydrogen and energy to the containment before the spray starts. Thus, the complexities of the spray's action are avoided. Conservative because the cooling action of the spray is avoided.

Thus, the pressure is maxinized. This arbitrary 15-minute period must not be used to criticize the starting time of the spray. It is more reasonable to consider Figure 2 of the submittal and observe that the hydrogen and

g 15 l

III. Accident Analysis (Contd) energy are actually released over 2 30-minute period. The spray will be on fo~ the last 15 minutes of this period. For this reason', the 15-minute delay in the spray initiation is reasonable.

It should also be remembered that the automatic initiation of the building sprays is a bsekup and that the operator can initiate the sprays. earlier. This he would be expected to do in the event of a major j

system rupture as soon as he had checked to see that the react was properly shut down and being cooled by the core spray.

f 1

CONSUMERS POWER COMPAI 4/A kJ

/Tf Vice President k4 P"C { / 7, /fh Date:

Sworn and subscribed to before me this /7 day of March 1966.

[

s n G --(L~

N) /

i

~

"No'tary Public, Jackson County, Michi6en My commission expires February.2 4, /9e' f i

i 1

.I t

r I

l 1

,,,.2...-,__.~_-_--

...._.---,-_--_..._-..,..-_-.,-.j

s 16 e

REFERENCES 1.

Lustman, Benjamin: Zirconium-Water Reaction Data and Application to PWR Loss of Coolant Accident, Report No. WAPD-SC-Sh3, Bettis, Westinghouse (Pittsburgh, Pennsylvania, May 1957) 2.

Owens, J. I., Lockhart, R. W., Iltis, D. R. and Hikido, K. : Metal-Water Reactions VIII. Preliminary Considerations of the Effects of a firealoy-Water Reaction During Loss of Coolant Accident in a Nuclear Rea : tor, Report No. GEAP-3279, General Electric Co., APED, San Jose, California (September 1959) 3 Lemon, A. W., et al: Studies Relating to the Reaction Between Zirconium and Water, Repost No. BM1-115h, Battelle Memorial Institute, Columbus, Ohio (January 1957) h.

Furman, S. C. and McManus, P. A.: Metal-Water Reactions IX.

The Kinetics of Metal-Water Reactions. Feasibility Study of Some New Techniques, Report No. GEAP-3338, General Electric Co., VAL, Pleasanton, California 5*

Baker, Louis, Jr. and Ivins, R. 0. : A Calculation Demonstrating the Effect of the Zirconium Water Reaction on the Analysis of a Loss-of-Coolant Accident : Paper Presented at the International Symposium on Fission Product Eelease...., Oak Ridge, Tennessee, April 5-7, 1965 6.

Baker, L. Jr., and Ivins, R. 0.: Reactor Development Program Progress Report for December 196h, Report No. ANL-6997, Page 57 ( Argonne, Illinois, January 1965) 7 Coughlin, J. P. and King, E. G. :

J. Am. Chem Soc 72 2262 (1950) 8.

Lustman, Benjamin and Kerze, Frank, Jr. : The Metallurgy of Zirconium, McGraw-Hill (New York,1955) 9 Wicks, C. E. and Block, F. E.: Thermodynamic Properties of 65 Elements -

Their Oxides...., U. S. Bureau Mines Balletin 605, Washington, D. C.,1963 10.

Doug'. ass, D. L.: The Physical Metallurgy of Zirconium, Atomic Energy Review, Vol 1, Page 71 (IAEA, Vienna,1963) 11.

Dow Chemical Co. : JANAF Interim Thermochemical Tables, Midland, Michigan, 1960 to 1965 12.

Ryshkevitch, Eugene : Oxide Ceramics, Academic Press (Ncv York, 1960)

17.

REFERENCES (Coatd)'

13 Hansen, Max: Constitution of Binary Alloys, 2nd Ed. With Kurt Anderko, McGraw-Hill (New York, 1958)

14. Cottrell, W.

B., Coordinator: Nuclear Safety Program Semiannual Progress Report for Period Ending December 31,196k, Report No.

ORUL-3776, Oak Ridge National Laboratory, Oak Ridge, Tennessee (March 1965) b 1

4 1

l i^

4 l

L F

i I

l I

n

~

,e-

-,w--,-

7 7

y w.

,c.

,v--

-r.,

% m,-

w,w

--,,--, w

, 4,- w/

K Au... Jewsnmusm? poemveme oPreca: s ee4 - 7as e36 peOu, m MMEces's OATE OP DOCJutNii DATE eECElvfD NO g74 g4 m

& L g gtqp tyn entmO:

REPoti-OTHER Jackson, Michigan TO.

CaiG,.

C C, OTHER-DOan X

U C73 M 'd ACitON NECE55Any 0

CONCUkeENCE O

DATE AN$wf DED NO ACilON NECESSARY O COMMENT Q/

BY CLAS$18 Post OFFICE FILE COCL REG YO DESCRIPi!ON. ' pus' to Unclospaed)

REFERRED 10 DATE RE CElvEE. sy DATE Ltn substitting the following as a regd,y to our 1-24-66 ltr:

Boyd'-

'W lifJ all extmn 40R ACTI(.i

' ENCLO5vaES, (32 cys ree'd --notarized 3 17)

Dr.Doan(w/mcr)

Answers to Addt1 Info Begineated Re: Proposed Change Noe 8

/

85"^"* 5-Distribution:

1 - forinal filo 1-tile I 1-Poa v

r~

2 - Coundiance I...., r, - e.I f W

1 - Stoner

'-4 Lc

%0 1. arL*

u. s. 4towc tNenov comissioN MAIL CONTROL FORM roagr,c.32es l

l l

l t