ML19323F856
| ML19323F856 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 06/24/1979 |
| From: | Cohen P METROPOLITAN EDISON CO. |
| To: | |
| References | |
| NUDOCS 8005290549 | |
| Download: ML19323F856 (10) | |
Text
.
ga
& 2.?
ye, n y 0u n k in e. f 2:lM l~) 'y THREE MII.E ISLAND--2 ACCIDENT OF MARCH 28, 1979 0%YGEN GENERATION AND GAS COMPOSITION OF SU33LE 8
?. Cohen, Consultan:
i June 24, 1979 Y
L 1.
Sources of Oxygen--Upper.3ound Esti=aca r
0xygen en:ered :he Three Mile. Island-2 System from :wo sources.
.These are, (a) oxygen in: reduced with the injecrion 1
l wc ar, and (b) oxygen crea:ad by radiolysis of the wa:ar in the sya:am by the decay radiati.on energy of the cora.
.1 0xygen in Injection Watar I
is assumed censarvatively':har the injection. water was air sa:= rated a: ambian: :empera =es, and thus would have an oxygan centan: of approxima:aly 8 ppm.
I: is fur:har assumed tha:-
i d=ing :he inciden; 250.,'000 gallons of wa:ar were injected. nro the sys:am.
This calenla:as :o a =ctal volume from this so
.e of 130 f=3(S.T.P.).
1 ppm 22.400 e=/g mole
=
32,000 mg/g mole 0. 7== 0./kg 0.,
=
4 250,000' gal x 8 lb x 0.454 kg gal l's 908,000 kg g0
=
2 To:al f 3 cc e 8.0 pgmx 0. 7 2
=
qpm - xg 0 2 x 1 f:3 X 317 cc 180 f:3 02
=
=
8005290 M.p
._ _ =
t
~
2~
In view of the extent of radiolysis as a source, it is
~
not wor:hwhile :o a :empt to refine this estimate.
1.2 Radiolysis of Reac:or Coolant The mechanism of radiolysis of water by reactor radiation (gamma, b e ta, and neutrons) under non-boiling and boiling condi-
- ions is discussed'in detail in Chapter 5'of Ref. 1. In summary, under non-boiling conditions, radiolysis of pure water is not continuous with energy input., b'ut rarher achieves an e'quilibrium.
degree of radiolysis wtiich is proportional to the square root.of
.the anergy deposition rate.
The equilibrium value is increas.ed-if the water contains an initial excess of oxidizing species (oxygen, hydrogen peroxide) and is decrea' sed if the water contains
\\
an initial excess of reducing species (hydrogen).
The latter n
condition is the preferred method of opera ion 'of pressurized '
wa:er reactors, in which, by maintaining an excess of E2 in the water (nominally 25 ccH2/kg of water) radiolysis is effectively suppressed in the reactor opera:ing at full power.
Under vigorously boiling (or bubbling) conditions radiolysis is a continuous function of energy input.
.The amount of gas formed is 0.45 molecules of H2 and 0.225 molecules of.02 PS:
100 ev of energy absorbed under chase condi: ions, From the decay energy curve for the reactor it is therefore possible :o calculate the." maximum" amount of oxygen which could have been generated from the energy available if all the other requirements had been met.
The "=axim:m" or 'uppe; bound yield per W4-hr of decay.
energy available from the core is calcula:ed as follows.
The c
i 3
decay energy is approximately equally divided between ga=ma photons wi:h an average energy of 0.7 Mev and be:a par:icles wi:h an average energy of 0.4 Mev.Ref. 2 The beca par:icles have a limi:ed range and for low enrichment fuels (UO2) are not E*f* 1
~
s ' 6"-.can -.
The gama photons have a relatively long range and are absorbed by all':he materials in. :he core, in proportion to the masses exposed.
The fraction of the gama energy absorbed in the. water is therefore conserva:ively~ estimated as follows, asstz:ing further that all of 'the gama energy released fr'om the fuel is absorbed in the core ' region.
Mass fraction of water in core re'gion, equal to fraction of gama energy absorbed in water in core is calculated from the R#
fuel assembly volume fractions,e*
- 3 and the material densities at temperature as shown below.
Fuel 0.303 x 10.00 =
3.030 0.406 Wa:er 0.580 x 0.7
=
0.653 Zircoloy 0.102 x 6.4
=
4.089 Sum
=
0.406/ 4,089 Mass fraction of water
=
0.10
=
"*he otal radiolyric energy "(R.E.) available for radiolysis in water per MW of decay energy is therefore R.E.
= (0.1) x gama + 0.0 x beta 0.1 x 0.5
=
0.05 %'RI/MW Decay
=
o l
e
e g-The yield of gases from radiolysis, per MW hr of decay i
~
\\
energy _s new calcula:ed as follows.
6 1 MW x 0.05 Mk'RE x 10 watts 87-MW x 3600 sees, 1.8 x 158 wat seconds
'r Radian: Energy n
MW hr
+ 1.593 x 10-19 wat: seconds elec ron volt
'+ 100 ev x 0.45 mol (H )
2 (100 e.)
(100,ev) v 24 5.08 x 10 molecules H /MW hr 2
23
/~
+ 6.02 x 10 mo'lecules
(
g mole x 22.400'ec/g mole 5
= 1.892 x 10 cc i
RU Er'
+ 28.317 cc/f:~3 6.68'f 3 H /MR hr
=
2 3
or 3.34 f:
0 /MW hr.
2 Table,1, shows the decay energy for TMI-2 as a fune: ion of ime af:er : shut-down, and the corresponding values of
" maxi =um" oxygen generation.
The decay energy values are taken from Ref 4 1
---__._v
~, _
s 5
TABLI 1 Time ?ariod, Average Max 0 LU *i"*
j 2
After Shut-down, hrs.
Decay Energy MW" Interval, ft f
' 2-3 26.06 87.0 l
3-4 23.00 76.8
)
4-6 21.07 140.7 6-8 19.13 127.8 8-10 17.74 118.5 l
l 10-12 16.63 110.1 12-14 15.25 101.9 1
76-2.8 Total /14 h=s
=
2.
Hydrogen Generation--Composition of' Gases.
The state of cooling of the core, during the course of the accident cannot be specified in sufficient detail to calculate
~
the actual radiolysis.
That part of the core immersed in water
- smains cool, and the water covering the core, which for the most part will be boiling, vill undergo radiolysis.
The amount of gas generated will be approximately that fraction.of the -
values of Table 1 corresponding to the fraction of the core covered by water.
The uncovered part of the core heats up rapidly to temperatures where the reaction of zircoloy cladding vi.th the steam generated f:cm the ' boiling section is quite rapid. Red. 5,6 The hydrogen released, and the decreased absorption in the low density steam result in low'radiolysis in that portion of'the core.
r
a,,
A 6
Essentially, the oxygen content of the gas released from the core was a maximum during the period when the I
system was boiling down, in the early part of the accident.
This 'is presumed 'to have oc=urred. when the last operating pump down at 100 minutes after the turbine trip.Re~#*7-was shut The state of the system, prior to uncovering of the core seems well described by the analysis of attachment 11, Ref. 5 which
~
illustrates the formation of steam' voids in the system.
- Actually, core damage may have started' earlier, as indicated.by a high.
reading on In-Core Thermocouple.10-R'at 32.5 tinutes. ef 7 R
Most of the gas generated in th'e early part of the' incident, largely hydrogen from the.zircoloy-water reaction; is pres'umed to have been vented to th'e' contnimnc.
This is
[
inferred from the evidence -for a hydrogen explosion in contain-(
i ment at 9 hrs 50 minutes.
It has been estimated enat 226 lb mol of hydrogen burned at that time.Ref h Ref 4, p. A12, A34, Attachent 3 Additionally it was estimated that 80 lb mol of hydrogen remained in containment, and that 76 lb mol of hydrogen was present in the. final bubble.
The latter figure apparently was escimated from the bubble volume, and the assumption, here believed to be erroneous, that the bubble gas.was a stoich-iometric =ixture of two volumes of hydrogen and one volume of oxygen.
This is supported by the following calculation.
3 Assume bubble volume was 10M f t at 250*E and 1000 psi Ref 8,9 The partia.1 pressure of j is 2.s 1.000 - 30 = 970 psi
(
(30 psia partial pressure o' u.v.;ei
,4 250 D.
The standard condition volume of gas is given by I
970 492 Y (ST?) = 1000 x x
B 77, E +250 45,720
=
H2(Ref. 4) = 76 x 359 = 27,284 V
VH2+02
=.1.5 x 27, 284 - 40,926 The values are close enough co. indicate that the.wiiter of Ref. 4 Attachment 11, did indeed assume that the bubble was composed of radiolytic gas, or had some high oxygen content.
The bubble could.culy have. arisen from compression of gas in the system above the reactor vessel after recovery and isolation of the system.~
This'probably included the following r
volumes (Ref. 3).
~
3 2 x 1/2 Steam Generator 2017 ft 2 Hot Legs 738 ft3 Upper Head of Vessel 508 fc3 3
2/3-of Upper Plenum 550 ft 1/2 of 2 Cold Legs 238 ft3 3
4061 ft During the entire period of preceding recovery, the s
void was sucerheated,to at least 700 F. Ref 8, Fig. 22.
The cold leg :e=perature varied, between a low of about 150 ? and about 200 ? just prior to recovery.
The pressure, Ref. 8, Fig. 5 was about 400 psi.
- Je can calculate the standard volume of gas in che voids as
+
e e
g x 788+460 = 46.857 f:3 40?
400 4060 x 1,,7 This is almost identical to the amount of gas in the final bubble.
The close agreemen: may,be fortuit'ous but the argument is believed to be sound.
Thus the oxygen conten: in the' bubble will not be greater than the average oxygen conten: for the maj or inciden: period plus some additional supply from radiolysis.
as the core was finally covered.
We esSimate that as follows:
Average composition of gas beford final covering:
2 = 3/4 of max radiolysis for 10 hrs (Table"1) 0 3/4 x 550.8 413 ft
/~
(
10/14 of 180 f:3 (injec: ion)
'129 f 3
=
3 542 ft H2 = T tal v lume fH2 150,780 f=3
= (226 + 80 x 114) lb mol
=
(x 359 f:3/lb col) 0.0036 0
=
2/H2 0, Oxygen addition on final covering, 2
3/4 of max radiolysis 10-14 hrs, Table 1 (212 x 0.75) 159 f=3
=
51 f 3 0
in inj ection wa.ter 2
210 f:3 3
4 180 ft x p-Ho
= Volume of bubble 45,000 f:3 0.0047 02
=
n2 Total 02
= 0.0083
= 0. 83 %.
r
9 l,
~'his is believed to be a conservative esti=a:e with respec to identified sources of oxygen and processes within the reactor system during the incident.
The true value is proba' ' y lower.
3.0 Su:m::ary and Conclusions It is evident that hydrogen re'ulting from the s
I zirconiu= water reaction was the major source of gas in. the j
i TMI-2 reactor system.
A minor quantity of oxyg'an may have' been introduced via the injection water.
Radiolysis of water was a minor source of additional hydrogen, and the probable source of most of the oxygen.
The total amount of oxygen available r
(
in the system was quite -small relative to the hydrogen.
The N-oxygen to hydrogen ratio in the gas was high only at the beginning of the accident, during the'first formation of a steam void in the system, before the core was uncovered, when the only significant source of gas.was radiolysis.
At this time however, the steam pressure was high, and the relative quantity of gas was low, rendering the mixture non-explosive.
Toward the,and of the incident, as the core was recovered, the relative oxygen content increased, into a high hydrogen conten void, but only to a final value which is estimated to be less :han one (17.), percent of the hydrogen, and thus' non-explosive at the existing conditions of approxi=a:ely 1000 psi and 250 y.
I l
10
~,
REFERENCES 1.
P. Cohen, "'4ater Coolant Tehenology of Power Reactors."
New York:
Gordon and Breach, 1969.
2.
S. Glasstone and M. Edlund, " Nuclear Reactor Theory."
D. Van Nostrand.Co., Inc., 6th Printing, 1957, p. 69.
3.
Three Mile Island-2' Data Package, Met. Ed.
4 NUREG-0557-Evaluation of Long-Term Post-Accident Cooling of Three-Mile Island Unit 2, May 1979, Fig. 6.2, pp. 6-16.
5.
G. ?. Marino, Preliminary Assessment of Core Damage fcir Three Mile Island Incident.
Memo for Files Fuel Behavior Research 3 ranch Division' of Reactor Safety Research._NRC April 25, 1979.
6.
M. L. Pickleseimer, " Bounding 'Es.d mates of the' Damage to Zircolay Fuel Rod Cladding in.the TMI-2 Core at Three Hours After the Start of the Accident," Fuel Behavior Research Branch NRC,' April 26., 1979; 7.
TMI-2 Interim Operational Sequence of Events as of May 8, 1979
(
EPRI.
. s.
8.
Preli=inary Annotated Sequence of Events, TMI-2 Accident 4
of March 28, 1979, G.P.U.
9.
PNO-79-67D.
O s
l l
~
4 l
l l
l l
i I
i 1