ML19331B828

Forwards Hydrogen Problems in Sequoyah Containment. Encl Is Second Part of Rept on Ice Condenser Plant Containment Response to Hydrogen Production & Burning & Mitigation by Igniters

ML19331B828
Person / Time
Site:	Sequoyah
Issue date:	08/04/1980
From:	Hubbard H R&D ASSOCIATES
To:	Gilinsky V NRC COMMISSION (OCM)
Shared Package
ML19331B827	List: ML19331B828
References
NUDOCS 8008130323
Download: ML19331B828 (16)

Text

--

• )

\_, ,

3D - 33 5' R & D ASSOCIATES Post Office Box 9695 lh Marina del Rey, California 90291 4 August 1980 Nuclear Regulatory Commission 1717 H Street, N.W.

Washington, D. C. 20555 Attention: Commissioner Victor Gilinsky

Dear Victor:

Enclosed is the second part of our report on ice condenser plant containment response to hydrogen production and burning and mitigation by igniters. If you have any ques-tions or comments, please call. We expect to see you and John Austin on Friday.

Best regards

'i Harmon W. Hubbard HWH/dl

Enclosure:

" Hydrogen Problems in Sequoyah Containment,"

August 1980.

1 l

, 8008130323 ,

a '

.. 6 - -

HYDROGEN PROBLEMS IN SEQUOYkN CONTAINMENT INTRODUCTION ,

This letter report completes the RDA response to a request from the Nuclear Regulatory Cohc.ission to critique the ultimate strength analysis of the'Sequoyah containment. This second report deals with the last two tasks of the work statement.

1. HGw would the an& lyses and results be altered if the stresses are caused by ignition / detonation of 300-600 Kg of hydrogen distributed uniformly and nonuniformly in the containment?

2. To what extenc can distributed ignition sources mitigate the effects of hydrogen?

A preliminary discussion of these topics was attended by Commissioner Gilinsky and Dr. John Austin at RDA on 18 July 1980.

l l

l 1

t O

seem w 4 - - , . . _ _ pv m- .h w==m m % - --* w -,y,#

pgeg, w w , ne _mmg

. e.

RESULTS

1. a) 300 kg of H2 gas mixed uniformly with the air and steam (if less than 40 percent steam) in the Sequoyah -

containment volume following an accident would be completely combustible if ignited (see Figure 1).

This complete co'bustion m could occur so rapidly as to exceed the capacity of the available heat removal proepsses, and could produce a pressure as high as 5.5 atmospheres, thus rupturing the containment (see Table 2). The combustion of 600 kg of H2 * "1d f course have more severe consequences.

b) A nonuniform distribution of 300 kg of H 2

present in the containment would consist of parcels of gas richer in H than the uniform distribution. If these 2

separated parcels formed while the blowers were operating, they would probably be mixed, combustible and perhaps detonable. If they were all detonable and all ignited, the damage to the containment would be worse than that due to ignition of a uniform mixture. If the gas parcels were not detonable, the pressure upon combustion wculd probably be at least as high as the uniform distribution. Under,some circumstances, it would be possible to collect pockets of gas too rich in H to burn. As the outer edges 2

• of such pockets mix with air, partially combustible mixtures would form.. The.results of igniting such a distribution would clearly depend on the sires of the parcels and the timing.

It should be noted that harmless mixtures of H 2

, airabd

. steam may become highly combustible or detonable as steam is condensed out (see Appendix B).

Thus one mechanism employed for removing heat from the containment also removes the combustion inhibitor from the containment.

2 O

a

2. If the rate of hydrogen formation is sufficiently low, and the mixing of H is complete and rapid so tnat all -

2 the gas in the containment gradually increases in H 2 .

concentration, then the presence of enough igniters could prevent overpressurizing the containment. This would be accomplished by releasing the heat of combustion at low concentrations over a long enough period of time to be handled by the heat removal equipment. However, if the 2r reaction rate is high relative to heat removal processes, then igniters might only delay containment failure.

Table 3 shows that a 1 percent per minute Zr re' action rate, accompanied by the burning of hydrogen at its rate of

, formation, would match the steady-state heat removal capacity of the RHR equipment.

j If the H is not thoroughly mixed, then there is a 2

! possibility of igniting a detonable pocket of gas with an l igniter. If left to its natural end, such an H -rich pocket 2

could disperse below the detonation limit- (420 percent H ) 2 when its ignition would cause less of a problem.

Since the possible rates of generation of H f 11 wiU9 2

an accident and the rate, place, and degree of mixing with

, air are highly uncertain, the use of igniters can only be an uncertain means of pressure control. Improper use might be detrimental rather than helpful. On the other hand, if it is assnted that there are many unavoidable igr.ition sources in the containment, it is certainly true that control of the time and place of ignition is preferable to chance. In this sense the use of igniters seems beneficial.

I COMMENT It is our opinion that the uncertainties in H2 generation ,

and mixing a,re so dependent on hardware details and scenarios that they are unlikely to be greatly reduced by further work.

For this reason we believe-it may be a better use of resources to explore thoroughly the feasibility of using an inert atmos-phere in the containment, so as to avoid the hydrogen burning problem. ,

. 3 i

u. .

d .

.i__.._.....~

~

s 100% air 300 kg -H 1000F saturated steam 2g i 600 kg H2 836 kg H2 \N '

150 F steam

\ x

\ ,200 0F steam Assumed detonation limit

\E -

w Flammability \

NNNNN N &

\

20% air

's /

100% H 2 3 # 100%

steam Limits of flammability and detonation based on Shapiro and Moffette WAPD-SC-545, as reproduced in WASH 1400.

Figure 1. Uniform mixtures in the Sequoyah containment vessel. l l

1 W

e 6

4

2 -

~

TABLE 1. INPUT DATA FOR SEQUC/AH PLANT

1. Free volume of containment vesse1I ") 3.2 x 104m3 -

Wefght of contained air at 27'C,1 atm. 3.7 x 104 kg Gram moles of air 1.3 x 10 6 Gram moles of oxygen .

2.7 x 10 5

2. Weight of zirconium in core (b) 1.9 x 104 kg Gram moles of zirconium 2.1 x 10 5

3. Yield of 100% 71rconium-water reaction Weight of hydrogen 836 kg, Gram moles of hydrogen 4.2 x 10 5 Heat of reaction (c) Zr + H 2O

, 1.1 x 10 II joules Heat of H 2 burn (d) (to form liquid H O) 1.2 x 10 ll joules '

2 Total heat of reactirsn + burn 2.3 x 10 Il joules

4. . Molar quantities and partial air pressure ,

of saturated steam in containment l At 100*F (38'C) vapor = 8.1 x 104 moles = 0.06 atm.

150*F(66*C) = 5.9 x 105moles = 0.25 atm. I 200*F(93*C) = 8.4 x 105moles = 0.78 atm.

NOTES:

(a) Sequoyah Nuclear Plant, Preliminary Safety Analysis Report (PSAR),

Table 5.2-1 gives the total containment active volume as 1,142,000 ft3 ,

comprised of 730,000 in the upper compartment, 125,000 in the ice compartment, and 287,000 in the lower compartment.

(b) ~ Sequoyah PSAR, Tabel 1.3-1, gives the clad weight as 41,993 lb.

(c) G. W. Keilholtz, ORNL-NSIC-120, Annotated Bibliography of Hydrogen Considerations in Light-Water Power Reactors, Feb.1976, Table 1 Heat of Reaction = 122 to 137 kcal/ mole Zr.

(d) Lewis and Von Elbe,

p. 685, 68.3 kcal/ mole H20.

5 I

- - . - y.-.-.- .- , .

TMM 2

• H 2 Quantity 300 kg 600 kg 836 kg .

1. Percent Zr Reaction -36% 72%

1Q0%

2. Id es H2 ~

1.5x10 5 3.10 5

4.2x10 5

3.

Partial Pressure 9 300'k (atmospheres) 0.12 0.23

' H 0.32

4. Molar Ratio h, Uniform Distribtion 0.11 0.23 0.32

5. Detonatable Mixture, no s(team presentD) or Combustibk (C)"

C D D

6. H2 Concentration Multiplier Required relative to uniform mixture a a) to reach detonation regime 2.0 1.0 1.0 b) to reach stoichiometric ratio of 0.42

1 for H2 : air 3.8 1.8 1.3

7. Steam Vapor Pressure Required:b a) to prevent detonation of unifonn mixture

. 0 0.1 atm 0.4 atm b) to prevent combustion of unifonn mixture 0.9 atm 2.0 atm 2.3 atm

8. Energy Release in 100% Combustion, Joules (liquid water product) 4.3x10 10 8.6x10 10 1.2x10 11 9.

Final Absolute Pressure in Adiabatic Combustion (Initial Air Partial Pressure 1 atm, Initial Temperature 300*k)c i

"a ) No steam,100% combustion 5.5 atm 10.0 13.3 atm b) No steam, 50% ccmbustion -

3.3 5.8 7.3 c) Steam 9 190*F, 50% combustion j 4.1 6.5 8.3 NOTES:. [

t t

(a) Approximate, based _ on regimes outlined in Figure 1. '

(b)

Approximate, based on regimes outlined in Figure 1, plus molar ccncen-trations of saturated steam c: a function of pressure.  !

(c) t Assuming p'roducts a constant-volume reaction' of combustion behave as ideal gases, and assumin i

t 6

t

• "~ .

es .

TABLE 3. HEATING'ANO COOLING RATES I SEQUOYAH CONTAINMENT Time when Fission Product Heat (Cumulative) Equals Total ,

Heat of Reaction 3000 see Rate of Heating at the 1% per min Zr Reaction Rate Zr Reaction 18.0 MW H2 Burning 20.0 Total 38.0 MW Rate of Fission Product Heating at 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> (when ice has been melted in DBA) 27 MW Steady-state, Cooling Capacity of the 2 RHR Heat Exchangers 67 MW Net Margin of Cooling Capacity (Beyond Chemical Reactions 01%/ min and Fission Product Heating) 2 MW NOTES:

(a) Sequoyah PSAR, Table Q.3-2 cites 2 heat exchangers, each having a capacity of 1.15 x 10o BTU /h at specified conditions.

I l

7

{ ,

j

c3 , ,

APPENDIX A LITERATURE SEARCH ON EXTENT OF HYDROGEN

• 1 BURNING AND FLAMMABILITY LIMITS FOR MIXTURES OF H , AIR, AND STEAM 2 l l

In considering the effects of 300 kg to 600 kg H 2 in the Sequoyah containment vessel, questions of lean mixture flamm-ability' limits and the extent of combustion are important.

The 1976 litkrature survey by Keilholtz (1) provided citations for most of the sources used in this brief study, and provided much of the available data on flammability and extent of ~

l combustion.

EXTENT OF COMBUSTION .

Keilholtz states that combustion of 100 percent of the hydrogen will not occur until the hydrogen comprises about 10 vol percent of the H -air mixture. A partial combustion 2

, data point of 50 percent combustion is quotea for a 5.6 vol percent H 2 mixture in air. This point is attributed to Shapiro and Moffette (2), a reference that we were unable to obtain in the available time. However, Furno, et al. (8) indicate about 90 percent combustion for an initial mixture of 8.5 percent H as compared with 5-10 percent combustion

' 2 for mixtures of.6.9-7.4 percent H . If 300 kg H w'#8 2 2 uniformly distributed throughout the active volume of the Sequoyah Unit 1 containment vessel, it would constitute a 10 vol percent mixture with air (neglecting steam), and hence could burn completely.

FLAMMABILITY LIMIT i The lean mixture threshold of flammability is given.by Keilholtz as 4.1 vol percent H in air bu.t at this concentra-2 tion, Egerto,n (3) as well as Keilholtz point out'that~the flame front is not coherent, and flame propagation is upward only.

. A-1 we e 1p emme-_- ~**h+ '

-'N**""

M

Downward propagation begins with a hydrogen concentration of about 9 vol percent (1), (3). Drell and Belles (4) state

~

that a 9 percent mixture will burn completely (a point'to be >

compared with the Keilholtz 10 percent mixture for 100 percent combustion). Even the lean mixture non-coherent flames are postulated to burn a mixture that is richer than the original mixture, because the high diffusion rate of H2 permits access of additional H to the flame (4). The diffusion rate of H 2 2

is also important to the dispersal of segregated pockets of hydrogen, and will be discussed later.

STEAM DILUTION The effects of dilution by steam are potentially important.

Drell and Belles (4) state that inert diluents have scarcely any effect on the lean-mixture limit of flammability, where 300-600 kg of H in Sequoyah would be, if uniformly distri-2 buted. They claim water vapor has effects similar to CO 2'

and they show data of Coward and Jones (5) (which we were unable to obtain) such that only after more than half the mixture is CO d es the fraction of H required for flamm-2 2 ability begin to increase. These findings are consistent with the ternary mixture chart of Shapiro and Moffette for

, H2, air, and steam, wherein the lean mixture flammability limit is at a nearly constant H fraction as the steam content 2

increases from zero to about 50 vol percent.

DETONATION Shapiro and Moffette indicate a triangular shaped detona-tion regime in their ternary mixture chart, a regime bounded

, approximately by a 19 vol percent H 2 line at the lean mixt 6re boundary and a 45 vol percent air line at the rich mixture '

boundary. Although the original reference was not available to us, it appears that the authors constructed the detonation l

regime by extrapolating from data on dry mixtures of H 2 and A-2 l .

I

_, m._ ___m ___ __-__ ,

s. .

air. We note that Drell and Belles show the range of deton-ability of H in Air from 18 . 3 V 1 percent to 50 vol percent 2

H.2 We could find no information on the effects of inert -

diluents on the detonability of hydrogen-air mixtures, and we note the caption on the Shapiro-Moffette ternary mixture chart: " Assumed Detonation Limits." We conclude that the effects of steam on detonability of H -air mixtures are 2

essentially unknown. The nearest information we could find was cited by'Keilholtz, and this pertains to detonations in Knallgas-team mixtures (6). Knallgas is a stoichiometric mixture of H 2 and 0 .

2 In reference (6), experiment's indicated that a minimum of about 65 vol percent Knallgas in saturated steam at 100'C was required for detonation. This would correspond to about 44 percent H

• 2 The occurrence of detonation is also influenced by the size and config'uration of the vessel, and the nature of the walls (4,7), which further complicates efforts to predict detonation precisely.

O 4

e e

A-3 e

REFERENCES

- 1. G. W. Keilholtz, " Annotated Bibliography of Hydrogen -

Considerations in Light-Water Power Reactors, ORNL-NSIC-120, Feb. 1976.

2. Z. M. Shapiro and T.- R. Moffette, " Hydrogen Flammability Data and Application to PWR Loss-of-roolant Accident,"

WAPD-SC-545, 1957.

3. Sir Alfred C. Egerton, " Limits of Inflammability," Fourth Symposium (International) on Combustion, the Williams and Wilkins Co., Baltimore, 1953, pp. 4-13.

4. I. L. Drell and F. E. Belles, " Survey of Hydrogen Combus-tion Properties," National Advisory Committee on Aero-nautics Report 1383, 1958.

5. H. F. Coward and G. W. Jones, " Limits of Flammability 'of Gases and Vapors," O. S. Bureau of Mines, Bulletin 503, 1952.

6. J. A. Luker and E. C. Hobaica, "Effect of the Initial Mixture Density on the Formation of Detonation in Knallgas

~

Saturated with Water Vapor," Journal of Chemical and Engineering Data, 6, 2 April 1961, pp. 253-256.

i

7. L. B. Adler, E. C. Hobaica, and J. A. Luker, "The Ef'fect of External Factors on the Formation of Detonation in Saturated Knallgas-Steam Mixtures," Combustion and Flame, 3, 481, 1959.

8. A. - L. Furno, E. 3. Cook, J. 11. Kuchta, and D. S. Burgess, "Some Observations on Near-Limit Flames," Thirteenth Symposium (International) on Combustion, 1971.

l l

, A-4

. w .m n ~.s~ f mmn- ~r-- -~

( .

w- -

APPENDIX B HYDROGEN-AIR MIXING BY FAN ,

i Air recirculation fans are provided in the Sequoyah con-tainment for returning air to the lower compartment after a postulated blowdown. Two such fans are provided, each having a rated capacity of about 40,000'cim. The purpose of the fan-induced cecirculation is to convey steam produced by residual heating to the ice condenser, if the emergency core cooling system should fail (failure of the ECCS is also a situation that could permit a zirconium-water reaction and hydrogen generation). The design basis for the recirculation system is an air flow rate of 40,000 cfm, corresponding to the cperatica. of one fan. Some parataeters related to mixing and burning of hydrogen in an air flow of 40,000 cfm have been calculated, and are presented in Table 4. t The air velocities in the ice condenser and upper plenum are low. Nevertheless, the flow would be turbulent in the upper plenum of the ice compartment, so the flow entering the 1 upper compartment should be well mixed. If hydrogen were being generated by a 1% per minute reaction of zirconium (as an example), the rate of hydrogen flow would be about 10% of the air flow, giving a mixture containing about 9% H2. This would be combustible, according to the literature cited else-where in this report.

A reference calculation is illustrated in Figure 2, where mixtures of 40,000 cfm air and the hydrogen yields of various rates of zirconium reaction are plotted on the ternary mixture chart. Each reaction rate corresponds to a straight-line ' locus, with steam rate determining the position on any line. The one point plotted on 'each line is for a steam rate that corresponds to the heat release rate o.f the Zr-H 2O reaction and the latent

, B-1 o .

l l

.__.___ __ ~

~^ -

~ .

e

. heat of vaporization of water. It can be seen in Figure 2 that the yield of 2r-H 2 O reaction rates in excess of 26 per .

minute can produce detonable mixtures with 40,000 cfm of air if the steam content is sufficiently low. Rates of several percent per minute were calculated for some accident scen-arios in WASH 1400.

e 0

l l

1 l

B-2 S

Table 4. Air Circulation Parameters Design Data From Sequoyah PSAR .,

Number of Blowers 2 Capacity of Each Blower 40,000 cfm Ice Condenser: Flow Area (net) 1,326 ft 2 Height 49 ft Annular Thickness 11 ft Effective Circumferential Length 267 ft Lower Compartment Active Volume 2.87x10 5 ft 3 Total Containment Active Volume 1.24x10 6 ft 3 Derived Parameters, for One Blower Operating Air Velocity: a) In Ice Bed 30 ft/ min b) In Upper Plenum of Ice Compartment 14 ft/ min Air Reynolds Number in Upper Plenum 2.6x10 4

air (kinematicviscosityog/

9 50*C = 1.15x10-2 ft min)

, Air Residence Time in: Ice Compartment 1.6 min

Lower Compartment 7.2 min

.  : Total Active Volume 31 min of Containment i l l

B-3 e

a. .

l L

l 1

- - . _ - . - . . -. - - - ~ ~ _ _ . . - - - - -

s

m .. .

100% air .

\\ - Zirconium reaction

\' rates, percent / min

\

X / .5 0 N ,N x 1.0 Assumed -

detonation

\ \

limit N \ ,

h, 2.0 N

Flammability N

N 0.0 20% air N 4\

e & \\s 3 p s

, 100% H2 100%

steam Limits of flammability and detonation based on Shapiro and Moffette WAPD-SC-545, as reproduced .in WASH 1400.

Figure 2. Locus of state points for mixtures of 40,000 cfm air with the hydrogen yield of various zr reaction rates.  ;

l B-4 l .

,