ML20054E109
| ML20054E109 | |
| Person / Time | |
|---|---|
| Site: | Millstone  |
| Issue date: | 04/30/1982 |
| From: | NORTHEAST NUCLEAR ENERGY CO. |
| To: | |
| Shared Package | |
| ML20054E107 | List: |
| References | |
| NUDOCS 8204260145 | |
| Download: ML20054E109 (40) | |
Text
.
a Docket No. 50-245 ATTACHMENT NO. 1 Millstone Nuclear Power Station, Unit No. 1 Final Rule on Interim Requirements Related to Hydrogen Control April, 1982 82 04 2 6 0 Mf
o o-Preliminary Evaluation of Hydrogen Flamability at the Millstone Unit 1 Boiling Water Reactor ABSTRACT In response to the December 2, 1981 final rule on " Interim Requirements Related to Hydrogen Control " NUSCO on behalf of NNECO tas conducted an engineering evaluation of Hydrogen burn scenarios at the Millstone Unit 1 Station. The purpose of this evaluation was to detemine the specific physical conditions necessary to yield flammable gas mixtures within the containment volume, and to lay the groundwork for quantitatively assessing the likelihood of reaching these conditions.
For the inerted containment, using plant specific design data and existing flamability limit data for ternary mixtures (H, 0, N ), the following results have been 2
2 2
obtained:
(a) Zr-H 0 reactions alone cannot yield flammable mixtures unless an 7
additional oxygen source is present.
(Hence long tenn 0 control 2
is the dominant problem.)
(b) The dominant source for 0 is control air leakage from the M.S.I.V.s 2
and S/RV valves. This 0 source can be eliminated by plant modi-2 fications.
(c) Assuming that plant modifications to the control air system are accomplished, the most direct pathway to reach flamability is to assume a Zr-H 0 reaction in the range of 1-5% in conjunction with 9
the radiolytic decomposition (with complete stripping from solution and transport to the containment) of greater than 400 lbs of H 0.
2 Based on preliminary analysis it is concluded that radiolytic decomposition of this magnitude is not achievable.
Further analysis is defined which will confirm this preliminary finding.
(d) Very e) tensive metal-water reactions (even up to a 100% value) greatly reduce flamability concerns because the excess H concen-7 trations proportionally reduce the 0 content and actually yield more inert conditions in the contain$ent.
o 4 I.
Flammability Limits and Containment Volume Effects In order to assess scenarios where flamable gas mixtures are developed within containment it is first necessary to define what concentrations are needed to enter a flamable mixture region.
Published values obtained by Coward and Jones, and reported in U.S.
Bureau of Mines Bulletin No. 305 (Reference 1) were utilized.
These flammability limits are shown in Figure 1.
No credit was taken for the effects of self-inerting due to the appreciable amount of steam anticipated to exist within containment in the post-LOCA containment environment.
(Specific assumptions of steam content would of course be scenario dependent.)
Complete mixing of evolved gases in containment is assumed (mixing of drywell and torus gases) due to the normal operation of the drywell to torus downcomer during the blowdown, and subsequent operation of the torus to drywell vacuum breaker valves.
The Millstone Unit No. 1 plant specific free air volumes are as follows:
3 Drywell:
146,900 ft 3
Torus *:
108,200 ft (Ref.2) 110,600 ft
- The gas space volume in the Torus Region is water level dependent.
Technical Specification limits exist for allowable water level.
3-The number of liters of gas contained within the drywell can be 3
obtained by multiplication of the volume in ft3 liters /ft by 28.316 This yields free air volumes within the primary containment between:
6 7.223 x 10 liters V
= 7.291 x 10 liters containment For conservatism, the smaller volume is assumed.
Utilizing the Ideal Gas Law, with initial containment pressure near 1.0 atmosphere and temperature near 325 K, the total. initial number J
of moles of gas will be:
=PY 5
(1) n containment = 2.7085 x 10 moles gas RT Normal (non-inerted) air contains the following volume concentrations of gases (Reference 3),
i N
78.084%
l 2
l 0
20.946%
2 C0 0.033%
2 Ar 0.943%
(
H 0.00005%
2 l
[
iwa
.. While under normal operations the containment gas compositicn is required
- to have an 0 content less than 5%. Procedurally, however, 2
0 content is limited to below 4%. Plant modification, planned to 2
eliminate the air source from the MSIV and S/RY, as discussed in Section II, should allow holding 0 concentrations at or slightly 2
below 2%.
In this analysis it is assumed that the 0 concentration 2
is initially at 4%.
Utilizing these assumptions with the containment initially inerted, the 0 content must increase from 4% to 6% simultaneously with an 2
increase in H content from -0% to 4% to reach flamable levels.
2 Using the concentration limits of flammability a set of simultaneous equations can be written defining the critical number of moles of 0 and H which are needed to yield a flamable gas mixture.
2 2
3 "H
2 (2a) 2
= 0.04
[H ] + [0 ] + [N ] + [X]
"H
- "0
+ "N
+ "x 2
2 2
2 2
2 3
"0 2
(2b) 2
= 0.06
[H ] + [0 ] + [N ] + [X]
"H
+ "0 N + "x
+
2 2
2 2
2 2
where [X] is the concentration of C0, Ar, and all other 2
rare gases found in air.
By definition, if the containment is initially inerted:
(3) nN + "x = 0.96 (# moles initially in contai ment) 2 5
= 0.96 (2.7085 x 10 moles) a 2.6 x 105,,),3
, (4a) n
=
0.06 2
0.94 ("H + "N *"x) 0 2
2 0 + "N +"*)
5 9T (n (4b) n
=
0.04 H
2 2
2 Making substitutions and solving for n and n yields:
0 g
2 2
4 1.73 x 10 moles n
=
02 2
1.15 x 10 gg),3 n
=
H2 The initial number of moles of gases in the containment would be:
5 4
0.04 (2.7085 x 10 moles) = 1.08 x 10 moles n
=
02 ng 2
Hence, the minimal path into the flammable regime requires generation of roughly:
3 l
6.5 x 10 moles of 02 gas 4
1.16 x 10 moles of H gas 2
These values are slightly conservative due to the fact that concentrations just within the flamable region would require an ignition source to generate a deflagration.
~
. II. Hydrogen and Oxygen Sources When the reactor core is inadequately cooled for sustained periods of time, the Zircalloy cladding in the fuel assemblies will heat up creating the potential for a high temperature Zr-H O reaction. As 2
this reaction proceeds, H2 gas is generated according to the following chemical reaction:
(5)
Zr + 2 H 0 --+ Zr 0
+ 2H2+
0 2
2 For each mole of Zr reacted, two moles of H2 gas are generated. As previously noted, when the containment is initially inerted the addition of H alone is insufficient to obtain a flammable mixture -
2 thus potential sources of 0 must be considered.
2 NUSCO has evaluated the nature and magnitude of several existing oxygen sources and has concluded that the only two sources are (in the order of their significance):
(i) Leakage of control air through the air operators of the M.S.I.V.s and S/RV valves.
(ii) Post accident radiolytic decomposition of reactor coolant.
Based on existing plant operating experience it is known that for the existing control air system with only 2-3 days leakage 02 concentrations can increase from a nominal 2% value up to the l
l Technical Specification limits - thus requiring the addition of l
more N. Clearly this source constitutes the dominant 02 source 2
within the Millstone Unit No. 1 containment.
l
. The second source of 0 is the radiolytic decomposition of H 0 2
2 during the boiling phase of post accident recovery (i.e. the time between initiation of the LOCA and the time when decay heat removal in the core yields a subcooled liquid inventory).
It should be pointed out that radiolysis of H O produces H as well as 02 gas 2
2 (along with a number of other intermediary compounds and gases such as H 0 )*
22 To address the full spectrum of potential H /02 generation sources 2
the following process was employed:
(i) The extent of the metal-water reaction was assumed to be para-metrically decoupled from the amount of radiolysis.
(ii) The extent of the metal-water reaction was parametrically varied throughout all ranges (i.e. 0-100%) and the resultant effects on H /0 concentrations were c.:mpared to the known 2 2 limits of flammability within the Millstone Unit No.1 contain-ment.
(This eliminates dependence on. knowledge of the extent l
of the metal-water reaction).
1 (iii) For various extents of metal-water reactions the extent of the radiolysis decomposition was also varied parametrically in tems I
of the number of lbs. of H O decomposed (assuming the net 2
reaction: 2H 0.-*2H2 + 0 ).
Again, the resultant H /02 concen-2 2
2 trations were compared to the known flamability for the Millstone Unit No. 1 containment.
I t
. (iv) By reviewing the resultant curves, the regions requiring the minimum extents of metal-water reactions and radiolytic decom-position were then identified.
Completion of this analysis identified the areas of greatest concern for achieving flammability.
The parametric evaluation of the extent of metal-water reaction (step ii) utilized the following information. The total volume of Zirconium in the cladding of the Millstone Unit No.1 reactor core is roughly:
5 3
(6)
Vclad = 2.45 x 10 in (Reference 4)
Assuming that all the zircalloy is zirconium (Reference 5 notes that zircalloy is > 98.5% zirconium) this would be roughly:
fZr (7)
M
- Y Zr clad Utilizing an atomic weight of 91.22 grams / mole the total number of moles of Zirconium in the clad would be:
"Zr 3.529 x 10' moles (8) n
=
=
Zr 91.22 gr/ mole i
l t
. Recalling that the reaction of one mole of Zr yields two moles of H2 gas, the number of moles of H2 produced may be expressed:
i (9)
AM b
O H.
Ar - Ha,o le 2
where is the extent of the metal-water reaction.
Values of
(,
range fiom 0.0-1.0.
Assuming only the effects of the metal-water reaction, the relative concentrations of H /02 gases may be expressed:
2 H l** -
(10a)
=
2 2f,,..
o
- An Hr.4 0*
N.+ n3 n,+n,+ng+ n e
o x
2 2
00, no7 2
0.d mgngnen~ g (10b)
,2 x
j gg Figure 2 shows the results of possible H2 generation scenarios (for metal-water reactions alone), with respect to the flammability limits.
Having assessed the effects of various levels of metal-water reactions, the assessment of the effects of various levels of radiolytic decom-position (stepili) was then carried out. Assuming for the present that net radiolytic decomposition proceeds in the forward direction *(the complete stripping of all dissolved gases is assumed with no recombinant back reactions), the net reaction would be:
- The evaluation of the likelihood of this occurring is assessed in Section III.
. n,p,Y (11)
- 2. H 0 v 2Hz+0 z
2 One mole of H O weighs:
2 (12) mH 0 = [2(1.0080 grams / mole) + (15.9994 grams / mole)] 2.20462 lbs 2
3 10 grams
= 0.0397 lbs. H O 2
Thus, each pound of H 0 decomposed, yields:
2 1 lb. H O --+ 25.178 moles H gas 2
12.589 moles 0 gas assuming a complete stripping of the dissolved gases and ultimate transport to the containment environment.
The number of moles of H /0 produced as a function of the number of 2 2 pounds of H 0 decomposed may then be expressed:
p (13a)
A ngl = E,o X w,,
Xyz= E A MG u
E ggo Xo,,
Kg: It.5M rwoles (13b)
AM 5
o g
b is the number of pounds of H O decomposed via radiolysis.
yg 2
I 1
I
. Using these expressions, the relative concentrations of H /02 gases 2
can be expressed as follows:
Ang,4 on,
w (14a) :.g2. rel.,
an d an/,4n,+ano,, + nw,+ nx a
o 2f -H,000r 4 Euf Xw,.
=
fr n,+ E
. (Yu,+ X.,.) + no,.+nye o 24.a,o e
S x
(14b) nog 4 ANos 1-rel.
A n,+ an ',4 n,+4n, + o / o a
u o
o w
x Oog4 b,o Oz H
g 2 fe,..g, n, + E u, (X H,.+ Xo, ) + n,+ n,+ n e
o u
x Figure 3 shows the relative H /0 concentrations with respect to 2 2 the flamability limits for[g,_g,*.0l(design basis LOCAffr-We=.0095) g andwithEg,o varied from 0 to 500 lbs. As is noted, flamability will be achieved if there is radiolytic composition of roughly 500 lbs.
of H 0.
2 Figure 4 shows the relative H /0 concentrations (with respect to 2 2 the flammability limits) for (,,g".05 andwithEu,o varied from 0 to 600 lbs.
Figure 5 shows the relative H /0 concentrations (withrespectto 2 2 theflammabilitylimits)forf,,,4, 5.10 and with E o varied Ha from 0 to 1400 lbs.
$ Qualitatively, these results indicate the following:
(i) The most critical region to be concerned about flammability is for[te-Hps.05 and En,o between soo-soo ibs. This region corresponds to the minimal amounts of metal-water reaction and radiolysis for which flamability can be achieved.
(ii) When the containment is inerted, the need to assess core degradation uptothedge.Ha0 35 evel, is unnecessary in that the required l
amount of 0 from radiolytic decomposition would be beyond credi-2 bility.
Having defined the region of greatest potential for becoming flammable, it is ncw necessary to assess whether the radiolytic decomposition in the range of 500-600 lbs. of H O is in fact possible.
2
III. Preliminary Assessment of the Radiolytic Decomposition Potential The initial decomposition of water occurs when a water molecule interacts with an energetic OL, h,Y,orneutron.
The net end result is the generation of Compton electrons as energy is transferred and absorbed by the water molecules.
Neutron effects are known to dominate during normal power operation and for the first several hundred seconds following reactor shutdown.
Following normal reactor shutdown the Y-dose would tend to become the dominant effect with a lesser contribution from fparticles (the majority of which are stopped by the fuel cladding). Where fuel failures are postulated to exist, the effects of ot and h type radiations can become larger due to the presence of dissolved fission products in the water.
In view of the fact g(H ) f r M particles is roughly four 2
times the value for mixed f,T radiations, and that c(-particles are roughly 1200 times more ionizing, one might expect the results for cases where extensive fuel damage has occurred to produce extensively more H and 0. Haissinsky, however, notes in Reference 15 that 2
2 0(-particles do not show significant differences in terms of net yieldsfromthoseexhibitedbyTrays.
The most likely explanation for this effect lies in the fact that, because or, particles are so densely ionizing and thus leave such short ionization' tracks (or
" spurs"), the recombination effects are heavily effected by charged particles diffusing over very short distances.
Practical experience in this area lies with the nonnal operation of PWRs which use soluble Boron-10 as a reactivity shim.
There are several models which can be utilized to assest the feasi-bility of decomposing 500-600 lbs. of'H 0.
In the discussion which 2
follows decomposition of 400 lbs. H O will be used as a conservative 2
upper limit. All of these models currently present certain limita-i tions in their usefulness and these are discussed below.
4 j
The Nuclear Regulatory Commission (NRC) has made extensive use'of the COGAP model (C_ontainment G_as A_nalysis Program) for assessing the contribution of radiolytic decomposition to the flamability potential in the post accident containment environment (References 6, 7).
This model utilizes a fission product decay curve and assumes a i
fraction of.the decay energy is absorbed in the coolant, thus producing free H and 0 m lecules, which are assumed to migrate 2
2 I
into the containment atmosphere. This model is adequate during the post accident phases where bulk boiling is occurring if the bulk yields (orGvalues)ofH/0 are correctly chosen. The COGAP 2 2 l
model, however,- is not appropriate when subcooling is restored in t
the core region. COGAP does not account for the important recombination 1
(or back) reactions which predominate in the subcooled phase and which would tend to halt any net production of H /02 gases.
(Asan 2
i example,:COGAP would predict an unending build-up of H /02 gases in 2
the reactor coolant system of a pressurized water reactor even i
during a normal plant shutdown). Additionally, the C0 GAP model will predict an appreciable unending net production of H /0 I"
2 2 the torus region even though no boiling is occurring.
i i
It should be pointed out that the NRC's use of COGAP to model gas generation following the accident at TMI-2 directly lead to the 1
l erroneous conclusion that 02 was building up in the steam bubble i._.
4 above the core region. Both the NRC Staff and the Kemeny Commission reviewed the NRC's assessment of_the hydrogen bubble and concluded that recombination effects should not have been disregarded (Ref-erence8, Reference 9). Because of the inability of the COGAP model (and Regulatory Guide 1.7) to accurately predict the known recombination effects, it is inappropriate to use this model after subcooling is achieved.
Schmidt of Argonne National Laboratory (References 10 and 11) developed numerical simulation models of water radiolysis and recombination in the late 1960's. This work and subsequent refine-ments lead to the WR-20 code (currently on file at Argonne National Laboratory). The WR-20 code models both radiolysis and recombina-tion effects via simultaneous solution of all the dominant first and second order homogeneous chemical reaction kinetics equation.
The code utilizes the absorbed energy dose rates in water, initial molecular, ion, and radical yields (or g values), and the chemical reaction rate constants; the code has been successfully benchmarked.
. The WR-20 model, however, is inappropriate for predicting radiolysis/
recombination effects when bulk boiling is occurring in the core region. Because of this, its usefulness must be limited to the regime where subcooling has been restored in the core region.
In view of these limitations, the approach that NUSCO has utilized to assess the possibility of decomposing 400 lbs of water is as follows:
(1) Utilize the COGAP model for core radiolysis during the first 4 to 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> during the boiling phase of the design basis LOCA.
Complete stripping and transport of all evolved gases to the containment is assumed.
(ii) Utilize the WR-20 model for radiolysis in the reactor drywell sump and torus regions.
(This is appropriate due to the fact there is no bulk boiling in these regions and hence recorrbination effects should dominate the results.)
(iii) Utilize the WR-20 model for radiolysis in the reactor core region for time periods after bulk boiling ceases.
For step (1), the mass generation rate terms (1bs/sec) for H /0 2 2 may be defined in a similar manner as'in COGAP (and Regulatory Guide 1.7).
Respectively these are:
. M mu (lbs./sec,}
N (15a)
H t
g l00 No 0
)
z z
(Ib5./'3cc,)
(15b) 2 100 No where:
P is the initial core thermal power, or 2051 MWt (the g
102% licensed power value)
G,,,,G.,are respectively the net radiolytic yields for "2 o2 in molecules predicted per 100eV absorbed dose.(hy :.M,bo,%
t Mug,Eoare the molecular weights of H, 0 respectively in grams 2
2 per mole. (mg : 2.016 F/vn,lc, m,= 31999 Wmole )
o g
E et) is the time dependent fission product energy absorbed in 3
the reactor coolant in units of eV/sec. MW
- t 0.0022 converts mass in grams to mass in lbs.
23 N is Avogadro's number or 6.023 x 10 molecules / mole.
g The rate at which water is decomposed in lbs/sec. by radiolysis during the post accident boiling phase may be expressed as:
e (16) bH oM N3
- d
%)
Nt t
~5 P. Ey3) h@
2.2 x 10
=
CT Mo
~[g,)
g, Hg Hz 0 2 g
. Substituting the appropriate values yields:
et)
kg,g=G.mGxlo'**Ey (37)
The decay energy (using data from Reference 6) may be expressed:
-x2t
( e.p
. sec.)
(18)
E.yli) = h K.
kr e gg i.
y is the fraction of energy absorbed by the where:
reactor coolant (for Millstone Unit lU.1: f '.05
)
y 4
k) = 5.1912 1 = 9.8 x 10 k = 0.8743 A = 6.5 x 10-6 2
2 4
k = 0.6557 A = 5.7 x 10 3
3 k = 0.4098 A4 = 7.4 x 10-8 4
k = 0.0150 h = 8.0 x 10-10 5
5 Using these expressions, the rate water is decomposed by radiolysis may be expressed:
y
)= 3.3733x 10 ] k; e
(
'/sec.)
d (19) i The net mass decomposed by radiolysis may be obtained by integrating the previous expression. This yields:
e
. 4 (20) N g o (Q go(h)ft = 3,3733)qo Q (g.d
) gg,s) g 2 Ac Utilizing this expression the following tabulated values are obtained for the net decomposition of water.
Net Mass of H O Decomposed 2
Time by Radiolysis I
1 x 10 sec.
0.0066 lbs.
2 1 x 10 sec.
0.24 lbs.
3 1 x 10 sec.
2.33 lbs.
4 1 x 10 sec.
17.66 lbs.
5 1 x 10 sec.
75.33 lbs.
5 2 x 10 sec. 121.0 lbs.
Figure 6 shows the time dependent decomposition of water. Also shown on this figure is the time period in which bulk boiling should cease within the core (i.e. 4-12 hrs. dependent on the particular scenario *).
It should be pointed out that this approach employs the following conservatisms:
(1) 102% of licensed power is assumed as an initial condition.
(ii) The decay energy curve is known to have a 20% conservatism in the long term.
(iii) Net yields of H /02 (Reg. Guide 1.7 values) are 10% higher 2
than published values.
3Design basis LOCAs will be evaluated for worst-case boiling interval and reported by NUSCO.
c (iv) No recombination is assumed at all.
In reality, vapor phase recombination will occur as well as an increasing liquid phase recombination as boiling intensity decreases in the cooler channels.
The calculations identify the fact that in the first stage (blowdown, refill, and boiling) less than 50 lbs. of H O will have been decomposed 2
assuming worst case radiolysis (e.g. no recombination is assumed).
The next area to be investigated is the subcooled stage. The radiolysis effects in the torus region (which can be assumed never to boil) and the core region following attainment of subcooled conditions, may be analytically treated by the same type of radiolysis/ recombination model.
In this regard it should be noted that the subcooled core water and I
torus water are directly coupled by the break path and the coolant injection path. For the purposes of chemical analysis, both may be assumed to have identical levels of dissolved fission products and impurities and be at roughly the same temperature region (for calcula-ting rate constants).
l l
While in the subcooled state there are a number of critical equilibrium l
effects which dominate the results. The coolant will decompose and recombine yielding the following net reaction as a result of a j
number of intermediaries.
nXp (21) 2 H O A 2 Hz + 02 2
l
o s The net buildup of equilibrium concentrations of [H ]eq and [0 ]eq 2
2 has been found to be sensitive to the following effects:
(a) the specific absorbed radiation dose intensity. Cohen in Reference 12 notes that all operating experience confirms the r
fact that [H 3 varies as the square root of the absorbed 2 eq radiation dose intensity.
Similar type effects would exist for [0 3 2 eq.
(b) the specific radiation type. All research in radiolytic decomposition / recombination has identified the differences resulting in initial yields (g values) due to radiation type.
Figure 7 tabulates the known g values for different radiation types.
(c) the concentration and types of impurities present in the liquid.
Jenks and Greiss (in Reference 13) demonstrated the effects of impurity oxidation / reduction type reactions on the net equilibrium values of [H 3 and [0 3 Additionally, Peled and Czapski 2 eq 2 eq.
noted changes in g values (Reference 14) due to presence of certain cation and anion solutions.
(d) the pH of the coolant. The pH of the water affects chemical kinetic rate constants and some researchers have noted (Reference
- 12) shifts in the initial g values.
1
e
-2 2-(e) the bulk coolant temperature. The bulk coolant temperature affects the chemical kinetic rate constants via Arrhenius' Law.
There are an appreciably large number of chemical reactions which can take place in the water following the initial decom-position of the water. These are summarized in Figure 8.
This list (for pure water) shows the chemical kinetic rate constants for near room temperature conditions. These rate constants can be adjusted to higher temperatures via Arrhenius' Law if the activation energy is known.
This list was generated in Reference 11. The equations listed are similar to those used by other researchers to simulate water radiolysis/ recombination effects (References 16 and 17). Specific values of the rate constants were chosen based on published data from the National Bureau of Standards (References 18, 19, 20, 21).
Using these reactions, it is noted that the kinetics of [H ] may 2
be described via the following rate equation:
d[Hal = K D (Hz) + k,[e.;] + k [e.(][H] + ki3[14]*
3 7
(22) d4
- k. [0H][H ]
2 K =(Avoytdro's No.)~'
D= absorbed p,Y olose in (" Miter.3.c.)
g(Hz) = inifial yield in molecules /ioo,y
e
. When chemical equilibrium is attained (dbt3 15 0) the H concen-2 tration in the water may be expressed:
K D g(Hz) + k, [eai]eg + k 7 [e.Q],3 [H]e3 + k,3 [ H],
(23) [Ng]eI
- ks[0H]eg i
From this expression it is noted that:
increasing radiation dose tends to increase [H ]q o
2 increasingconcentrationsof[Geh,pnd [H3e3 tend to o
increase [H ]eq 2
o increasing concentrations of [0H]
tends to decrease [H ] q 2
The chemical kinetics of dissolved 02 gas may be described by the following rate equation:
= ky [0H][0,-] + kzz [H0 ][0 -]- k,, [cag-][0 ]- k lH3@z]
2 2
2 iq 20 Whenchemicalequilibriumisattained(
- =0)the0 concen-2 I
tration in the water may be expressed:
+ k z [H0 ],f0 -],5 ky LOH],3 [0'],5 t
3 2
2
(
- '1-k,, [8ai3c3 + ki3 [H]cg l
From this expression it is noted that:
l
o o o 0 is not produced directly via radiolysis, but through 2
the subsequent reaction of ions and radicals.
increasing concentrations of [0H]eq, [H0 32 q, and [0 -3 o
2 eq tends to increase [0 ]eq.
2 increasingconcentrationsof(G, nd [H]
tends to o
decrease [0 3 2 eq.
The WR-20 Code (Reference 11) can be effectively utilized to solve for all of the key radical and ionic species found in the radiolysis/
recombination reactions.
One of the major effects which must also be accounted for is the reaction of metallic impurities with the key radicals and ions.
Qualitatively for a typical impurity species: Y,thefollowing types of oxidation / reduction reactions can take place:
y *
- 4 H --+ Y "~ + H +
l y " + e.y Y"~ + ( Hz0) f"+0;-+f*(**'+Oz y +(a-'i+ 0H --+
+ OH -
I Because of the fact certain impurities compete for OH, Egg ~, and H, their net effect is to change the. equilibrium concentrations of l
l H and 0 dissolved in the water.
2 2
1 i
-2 5-Clearly WR-20 cannot be utilized to predict the actual post-accident equilibrium concentrations, as there are far too many possible variations in initial conditions. What WR-20 can be utilized for is to define a parameter space of initial conditions (temperature, pH, impurity levels,' etc.) in which chemical equilibrium is predicted without any net generation of H and 0 gas.
Based on normal 2
2 operating experience and the experience from the accident at Three Mile Island (where there was extensive deposition of impurities in i,
the coolant) it is clear there is a wide range of conditions which preclude gas generation as long as the coolant is not boiling.
NUSCO intends to develop this infomation further and proper definition of the acceptable bounds where H /02 generation is precluded will 2
confirm the fact that decomposition of 400 lbs. of H 0 is not 2
credible.
4 Preliminary simulation of kinetic effects during radiolysis/ recombination reactions show equilibrium concentrations of H and 0 being achieved 2
2 in less than a second for the case of pure water. The equilibrium levels are found to be less than the solubility limits based on Henry's Law. Because of this, it is felt that the existing gas partial pressures in the containment may be the effect which dominates the equilibrium levels of H /0
- 2 2 An additional effect which may be significant in the long tem, is the gaseous phase H /0 recombination in the presence of significant 2 2 core and airborne activity. This should effectively serve to in the steam space above the core recombine a portion of the H /02 2
t e
m-.-
e
+w--ev-
-,ey
,en--=
,--wv.wv.
e_
--,.,.,av.
--,w.-
-y y
9- - _
-y
. region and perhaps in the torus as well. The rate of gaseous phase recombination in a gamma radiation field was measured by Benjamin and Isben in 1965 (Reference 22). Their work shows that the recombi-nation may be modeled via the following rate law:
~ a (N ),
k, D (Hg)(Oa)
(26) t dt I + k (0 )
t 3
k, = 8.4 x 10-2:
k = 2.04 x10~"
t D= absorbed dose in ("V[hr 3, Hzo)
(H ),(0 )=Sas partial pressures in mm H3 at 30*c 2
2 NUSCO intends to evaluate this effect as well.
i
o IV. Sumary and Conclusions The first portion of this report established the specific amounts of H and 0 necessary to achieve potentially flamable gas mixtures.
2 2
In the second portion of the report specific levels of core damage and radiolysis were defined which could lead towards these H2 and 0 levels. Additionally, the effects of N inerting were 2
2 established and it was shown that 0 control was the dominant 2
problem in avoiding flammability.
(In this regard the need for improvements in the present control air system were clearly iden-tified). The limiting area of flamability concern was identified as fuel damage between the 1% Appendix K limit and 5% - in conjunction with the net radiolytic decomposition of greater than 400 lbs. of H 0.
It was also pointed out that postulation of fuel damage 2
greater than this range yielded less flamable conditions.
The third portion of this report investigated the potential for decomposing greater than 400 lbs. of H 0.
In this regard it was 2
noted that the cessation of all bulk boiling within the reactor core region (12 hrs) puts an upper limit on the net radiolytic l
decomposition of less than 50 lbs. H 0.
Because of this, there is 2
apparently substantial margin in avoiding flammable conditions within the containment.
In order to confirm these preliminary findings, NUSCO intends to submit additional information further documenting the results of our analysis of radiolysis and recombination reactions.
It is
. intended that the following effects be examined for their effect on the equilibrium levels of H and 0 2
2 o
bulk water temperature o
metallic and fission product impurities o
kinetic effects of dissolved N, N0, N0 in the coolant 2
2 o
variations in pH o
rate of gas absorption into the water from the containment atmosphere
-o variation in containment gas partial pressures j
f f
~
REFERENCES 1.
Coward, H.
F., and Jones, G. W., " Limits of Flammability of Gases and Vapors", U. S. Bureau of Mines Bulletin No. 305, 1952.
2.
Millstone Nuclear Power Station, Unit 1, Technical Specifications.
3.
Handbook of Physics and Chemistry, Chemical Rubber Company 47th Edition, 1966-1967, p. F-123.
4.
Parker, D.
J., " Millstone Unit I Isolation Condenser Requirements for Station AC Blackout," NUSCO Calculation Wl-517-268-RE.
5.
Tong, L.
S., and Weissman, J.
E., " Thermal Analysis of Pressurized Water Reactors," Copyright 1970 by the American Nuclear Society, p.
66.
6.
Combustible Gas Control in Containment, Standard Review Plan Section 6.2.5, Rev. 02 issued by the Office of Nuclear Reactor Regulation, U.S.N.R.C., July 1981.
7.
Regulatory Guide 1.7, U.S.N.R.C., Revision 2, 1978.
8.
Technical Staff Analysis Report on Chemistry to the President's Commission on the Accident at Three Mile Island, October 31, 1979.
9.
"TMI-2 Lessons Learned Task Force Status Report and Short-Term Recommendations" NUREG-0578, issued July 1979.
- 10. Schmidt, K. H., "A Computer Program for the Kinetic Treatment of Radiation-Induced Simultaneous Chemical Reactions," Chemistry Division, Argonne National Laboratory, ANL-7199, issued April 1966,
- 11. Schmidt, K. H., "A Computer Program for the Kinetic Treatment of Radiation-Induced Simultaneous Chemical Reactions: A Revised Version in FORTRAN IV," Chemistry Division, Argonne National Laboratory, ANL-7693, issued March 1970.
l
- 12. Cohen, P., " Water Coolant Technology of Power Reactors, Gordon ar.d Breach, New York, 1969, p. 98.
- 13. Jenks, G.
H., and Greiss, J.
C., " Water Chemistry in Pressurized and Boiling Water Power Reactors," 0RNL-4173, issued November 1967.
- 14. Peled, E., and Czapski, G., " Studies on the Molecular Hydrogen Formation (GH2) in the Radiation Chemistry. of Aqueous Solutions,"
l Journ. Phys. Chem., Vol. 74, No.15,1970, p. 2903.
- 15. Haissinsky, M.,
Nuclear Chemistry and its Applications, Chapt.13,
- p. 385.
- 16. Burns, W. G., and Moore, P.
B., "A Survey of In-Reactor Zirconium Alloy Corrosion," AERE-R 8184, issued March 1976.
- 17. Burns, W. G., and Moore, P. B., " Radiation Enhancement of Zircaloy Corrosion in Boiling Water Systems: A Study of Simulated Radiation Chemical Kinetics," AERE, Publication E0-77-125.
- 18. " Selected Specific Rates of Reactions of Transients from Water in Aqueous Solution, Hydrated Electron Data," NSRDS-NBS-43, Issued May 1973.
19.
" Selected Specific Rates of Reactions of Transients from Water in Aqueous Solution, Hydrated Electron, Supplemental Data," NSRDS-NBS-43, Supplenent, Issued June 1975.
20.
" Selected Specific Rates of Reactions of Transients from Water in Aqueous Solution, Hydrogen Atom Data," NSRDS-NBS-51, Issued May 1975.
- 21. " Selected Specific Rates of Ractions of Transients from Water in Aqueous Solution, Hydroxyl Radical, Perhydroxyl Radical and Their Radical Ions," NSRDS-NBS-59, Issued January 1977.
- 22. Benjamin, B. M., and Isben, H.
S., " Recombination of Hydrogen and Oxygen in the Presence of Water Vapor Under the Influence of Radiation,"
Final Report AEC Contract AT(ll-1)-1032, issued July, 1965.
l l
l l
7.
e c
[21/I 1 + Ill ] + [N ] + [X]
2 2
2
.20
.18
.16
.14
.12
.10
.08
.06
.04
.02
.80 i
g i
i i
.70
,f10)%
d i 7 3%
.60 60%
507 e.
4 40%
.50
,i
/ 30%
.40 o
f 20%
[H l
.30 l
2
[0 )+(ll ]+[N ]+(X]
f 2
2 2
e
.20 p
10%
.10 45%
I e
S i
i e
i e
i 2 0%i
.10
.20
.30
.40
.50
.60
.70
.800.90 1.00 Additional Nitrogen Initially Added to Containment Air Figure 2 l
l l
l W
e e
i
[0 ]/[0 ]+[H ]+[N ]+[X]
2 2
2 2
80 20
.18
.16
.14
.12
.10
.08
.06
.04
.02 i
i i
i i
i i
i
{
.70 70%
[H l 2
.60 60%
j[0 ]+[H ]+N ]+[X]
2 2
2 e 50%
9
.50 4 402 I
0%
?
.40 l
f20%
.30 t
l
.20 f 10%
.10
$ 5%
2 I
'l I
s 20%i i
e i
e i
i 0
.10
.20
.30
.40
.50
.60
.70
.80
.90 1.00 Additional Nitrogen Initially Added to Containment Air i
l Flammability Points:
l l
(1) Concentrations resulting from
~1% Zr-H O Reaction 2
(2) Radiolysis of 500 lbs. H O 2
Figure 3
a.
c i
70 mo
&u 60
=w a<
M+
x, 50 cn
- O Carbon Dioxide 5"
40 Nitrogen ec o o
Stoichiometric 5*
30 Mixtures P
o oc 20 a8
~
~
57, v=
y 10 m
0 I
I I
I I
I I
l 0
16 14 12 10 8
6 4
2 0
Percent of Oxygen in Mixture l
of Diluent Gas + Air l
Fig. 1 Flammability Limits of flydrogen in Air Diluted with CO rN 2
2 i
e e
[0 ]/[0 ]+[H ]+[N ]+(X]
2 2
2 2
.20
.18
.16
.14
.12
.10
.08
.06
.04
.02
.80 l
i g
i i
i i
i
.70 fgg%
< > 80%
, 70%
60%
.60 e 50%
[H ]
.50 40%
2
-[0 ]+[H ]+[N ]+[X]
e' 30%
2 2
2
.40 l
9 4 20%
.30 l
l
.20 Flammable Region f10%
5%
.10 Il m
I I
I t
i e
i 1
i
.10
.20
.30
.40
.50
.60
.70
.80
.90 1.00 Additional Nitrogen initially Added to Containment Air 5% Zr-H O Reaction 2
(1) Radiolysis of l
500 lbs. H 0 2
(2) Radiolysis of 600 lbs. H O 2
Figure 4 l
e-
'e 1
l (0 ]/[0 l+III ]+[N ]+[X]
2 2
2 2
.06
.04
.02
.80 I
I I
t i
I I
i I
- f p 7C %
.60 f,60
[11,,)
E 50?
.50 4 40?
- [02 +III ]+[N ]+[X]
2 2
.40 o' 30%
l f20%
.30
+ 3 21
[
Flammable Region
'N,
/
.20
's '
j10%
j5%
.10 i
1 i
i 2
.10
.20
.30
.40
.50 60
.70
.80
.90 1.00 Additional Nitrogen Initially Added to Containment Air i
10% Zr-ll 0 Reaction 2
Figure 5 (1) Radiolysis of 500 lbs.
11 0 2
(2) Radiolysis of 600 lbs.
HO 2
(3) Radiolysis of 1000 lbs.
11 0 2
s '.
Figure 6 8-s 1
8 7
.o o
O 8
5x e
w g
=
e x
.O E
E it 8
~
O n
Uo i
e t-m E
o 3
R I,
I o
' Bulk
', Boiling l
Ceases g
I 1
l 8
I I
I I
I 1 hr.
i 12 hrs.
U-8 hrs.
o I
l I
4 10 10 10 10 2
3 Time in Seconds
Direct Yields from Radiolysis, g (
) in molecules 100eV
- a]
H+e, H
H N0 H0 2
22 2
L mixed'o',f 2.31 2.86
'0.55 0.44 0.70 2.34 m 0.00
.02 Y'
2.60 0.55 0.45 0.70 2.60 0.02
.02 R.
I 0.36 0.72 0.36
-1.12 1.00 0.47 0.17 4.0 eV R
10 (rt,a )7Li 0.04 0.20 0.16 1.70 1.30 0.10 0.30 24.0 eV B
R.
Figure 7
s x
t Figure 8 Chemical Reactions Occurring in Pure Water at Room Temperature 9
- . 20H" + H k - 5.5 x 10 e
-+e 2
l e - + 0H
% OH~
k = 3.0 x 10 2
9 OH + OH 22 3
H0 k = 4.5 x 10 9
OH + O~
02 + OH-k = 8.0 x 10 p
4
~
~
OH
+H e
+ (H O) k 2
5"
~ + 11+
H k = 2.4 x 1010 e
6 0
~
OH" + H k = 3.0 x 10 e
+H
.---p 2
7
~
~
e
+H0 OH + OH k
22
~ ~-.p 8
~
9 e
+ H0
+ OH 4 20H~
k = 3. 5 x 10 2
g
~
1 e
. - - - > H + OH" kl0 = 2.7 x 10 11 i
H + OH~
(H O) k
= 1.43 x 10 2
yy (H O)
---c.
OH~ + H+
k
= 2.59 x 10-5 2
g i
H+H p.
H k
2 13 10 11 + O H (H O) k
= 2.5.x 10 2
g 7
OH+H 2 g
H + (H O) k
= 4.5 x 10 2
15 7
011 + H 0 H02 + (II 0) k
= 4.5 x 10 22 2
16 H+H0 OH + (H O) k
= 4.4 x 10 22 2
y7
~
0 e
+O 0
k
= 1.9 x 10 2
2 18 0
H+O 2
= 2.0 x 10 2
19 1
H0
- . 0
+ H+
k
.0 x 10
=
2 2
20 H+ + O p H0 k
= 5.0 x 10 0 2
2 g
l-
a n
Figure ' 8 (cont 'd) k
= 7.9 x 10 H02+O2 y
02 +.H02 22 0
H + H0 -
p H0 kg = 2.0 x 10 2
22 0
~
H+O 11 0 k
= 2.0 x 10 2
2 g
=2.0x'10f0 a
+ HO H0 -
k
~
aq 2
2 25 e
+0 H0
+ OH~
k26. = 1.0 x 10
~
ag.
2 2
OH-'+H0 H0
+ (H O) k
= 1.0 x 10 22 2
2 27
~
~
k H0
+ (H O)
H022+OH 28
=
x 0 2
2 1
4 1
ll 4
i s
1 1
J w
c.
-, - --