ML19220B605

Forwards Brief Summary of Preliminary Conclusions of Probable Fuel Sys Damage Due to Incident

ML19220B605
Person / Time
Site:	Crane
Issue date:	04/06/1979
From:	Meyer R, Rubenstein L, Tokar M Office of Nuclear Reactor Regulation
To:	Case E Office of Nuclear Reactor Regulation
Shared Package
ML19220B586	List: ML19220B585 ML19220B601 ML19220B605 ML19220B606
References
NUDOCS 7904270090
Download: ML19220B605 (16)

Text

'. _

enarco4,

7 UNITED STATES p

NUCLEAR REGULATORY COMMISSION 3

f 4

S

$*k~,@,I [

April 6, 1979 WASHINGTO N. D. C. 20555 e,

'*S....'

[g h/

3

/

AHACHMENTD, MEMORANDUM FOR:

E. G. Case, Deputy Director Office of Nuclear Reactor Regulation FROM:

TMI Fuel Team

SUBJECT:

ESTIMATE OF FUEL DAMAGE IN THREE MILE ISLAND (TMI)

Enclosed is a brief report describing the preliminary ccnclusions of the team formed to analyze the probable damage to the fuel system at TMI.

[q ON u*

. C n1 Ar L. S. Rubenstein, PSS/NRR R.

Meye 7s. Th M. Tokar, D^S/NRR W. V. Johnston, RSR/RES W./

1

+

Enclosure -

As stated 7904'270090 80-242

9 On Tuesday, April 3,1979 a team consisting of L. S. Rubenstein, PSS/NRR R. O. Meyer, DSS /NRR M. Tokar, DSS /riRR W. V. Johnston, RSR/RES was formed to survey the fuel groups analyzing the damage to the fuel system of TMI and draw some preliminary conclusions from their deliberations regarding that damage.

The following individuals and organizations were contacted on April 3 and 4,1979:

E. L. Zebroski (EPRI) -

(representing the Metropolitan Edison Group)

J. Taylor /J. Tulenko, B&W R. Denning, BCL D. McCloskey, Sandia J. Scott, LASL In addition to the infomation obtained from conversations with these organizations and the NRC scaff, the team obtained a

" sequence of events" from B&W (Enclosure 1) a group of curves describing the pressure, temperature changes at TMI-2 during the first 15 hours1.736111e-4 days <br />0.00417 hours <br />2.480159e-5 weeks <br />5.7075e-6 months <br /> from D. Eisenhut, and a BAPL radiochemical analysis of the primary coolant taken at 1600 hours0.0185 days <br />0.444 hours <br />0.00265 weeks <br />6.088e-4 months <br /> March 29, 1979 and decay corrected to 0700 hours0.0081 days <br />0.194 hours <br />0.00116 weeks <br />2.6635e-4 months <br /> March 30, 1979.

The primary information used in our analysis of fuel system damage was obtained from the B&W Company, the Metropolitan Edison Industry Group, and from calculations of the NRC staff (Reactor Fuel Section, CPB; Fuel Behavior Branch, RES).

System Effects Using the chronology of events obtained f:om B&W and the control room strip chart tracing of system pressure for the first 15 hars of operation, we were able to ' determine that there were three periods in which the primary system pressure was below a saturation pressure corresponding to a temperature of 620*F. The system changes which caused these periods are described in i.he saquence of events provided by B&W enclosed with this report.

The details of what occurred to cause the pressure changes in the primary system are not f.iscussed here as these are considered in or.her staff reports (see e.g., IE Bulletin 79-05A, Nuclear Incident at Three Mile Island) and will be evaluated by others.80-243

A o

. Examination of figure 1 shows that the first period in which the system pressure was substantially below saturation pressure occurred The second approximately 1.75 to 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> af ter start of the transient.

period, which was relatively short in duration, occurred in the 4.5 to 5.5 hour5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> time frame and resulted in a small decrease in primary system pressure below saturation pressure. The final period of decreased primary system pressure extended from approximately 8-14 hours after start of the transient.

It was during these 3 periods that the core was exposed to extensive amounts of steam cooling and experienced fuel damage.

The group was able to infer from examination of these pressure histories, reports of fuel channel temperature changes with time obtained from the incore thermocouples, the behavior of the ir.. ore rhodium self-powered neutron detectors (SPND's), and 3'-long Intermediate Range Ex-Core Detectors, and the containment radiation monitors some details of when the fuel pins lost their integrity, the (?pth of the core which was exposed to steam cooling, the probable time periods of that exposure, and the amount of damage to the fuel.

As previously stated, the evidence for the leycl of uncove' ring was cbtained from a B&W analysis of the incore SPND's.

It can be shown that :

Above about 700*F, incore SPND's (Rh) act as thermionic elements and generate currents which are correlatable to temperature. Thus, if a discontinuity is observed in current measurement, a transition in temperature may be inferred.

It was assumed that this discontinuity rep-resents an elevation at which voiding of the coolant has occurred.

Similarly, the excore Intermediate Range Detectors may be used to provide an indication of voiding.

The information obtained from these detectors was consistant with the results from the Industry Group calculation that, in approximately one hour without introduction of makeup water, the core could boil down to full uncovering.80-244

Fuel System Conditions During Period of 1st Uncovering During the first period of major uncovering of the core (at least 5 feet of the core was uncovered for about an hour, and perhaps all of the core may have been uncovered for about one-half hour), the uncovered portion reached temperatures high enough to fail fuel rod cladding. At this pToint, fission products were released into the primary coolant ~ as evidenced by the subsequent alanning of the - - - -

containment activity monitors.

Based on the measured coolant activity and the amount of hydrogen release from reaction of the Zircaloy cladding with water, all of the fuel rods probably defected and released fission products.

. uel temperatures were estimated from calculations based on the fission product analysis of the sample of primary coolant, apd also from heat transfer considerations. Based on back-calculations that accounted for temperatures and temperature-dependent release rates that would be required to produce the measured level of activity, fuel temperatures of 1400 to greater than 1600*C were obtained.

Estimates by ORNL based on their experiments indicated that the Cs and I releases measured would have required fuel temperatures of at least 1300*C for an hour.

The heat transfer calculations indicated, on th3 other hand, that the fuel temperature may have been only about 1100*C.

In either case since is 2840*C, fuel melting was unlikely.

These the melting point of UO2 temperature differences can be rationalized by considering that a small There is portion of the core may have been at the higher temperatures.

and Zr0 also a possibility of ume eutectic formation between U0 at temperatures above approximately 1800*C, but no significance w$s attached tn the occur'ance of such a eutectic.

Later analysis by members

-ROM) indicates of ANS-5.4 f!ssion gas working group (including one ofy data and the fuel pellet tem >eratures as high as 2000*C based on Xe assumption that half of the core remained coal.

While noble gas activities lent. themselves to smaller analytical uncertainties than iodine or cesium activities, the uncertainty in the core fraction that is responsible te the release still renders this result inconclusive.

3 Hydrogen balance caiculations indicate thag from 15 to 30% of the total Zircaloy inventory has been oxidized.

Some of the oxidation, however, undoubtedly occurred during the latter uncoverings. The extent of the oxidation probably varies as a function of height in the core, with the greatest amount of oxidation having occurred in the uncovered (upper) portions of the fuel rods.

Later calculations accounting for hydrogen in the bubble, in the containment, lost in the hydrogen explosion, and gained by radiolysis suggosts that almost 40% of the Zircaloy in the fuel region may have been oxidized.

2 CPB Staff Calculation ICPB Staff Calculation Industry Group Calculation 3 B&W, Industry Group and NRC Staffs 4 Industry Group & NRC Staffs B&W Calculation 80-245

.s

. As the primary coolant level was restored during the latter portion of the time period of the first uncovering, thermal and mechanical shock loadings of the oxidized and embrittled cladding are believed to have occurred and to have resulted in cladding fragmenta

• ion.- -

At the end of the period of first uncovering, virtually all of the fuel rods had defected and released fission products. Al though temperatures had been high enough for a long enough time to have caused severe cladding oxidation, continued operation of incore instruments strongly indicates that fuel assembly structural members such as guida tubes remained intact.

Control rod materials are believed to have remained in place, as indicated by the absence of silver in the primary coolant.

Fuel System Conditions and Effects During Period of 2nd Major Uncovering At about 41/2 hours into the event, the core level again decreased to expose the upper 5 feet of the fuel assemblies.

The duration of this additional uncovering was shcrter than the first, the system pressure was higher, and the overall temperature effects were less severe, as evidenced by the fact that the thermocouples in the outer periphery of the core remained on-scale. Because of the reduced severity of the core conditions during the second uncovering, as compared with the first uncovering, less damage is believed to have occurred to the fuel system.

Fuel System Conditions During Period of 3rd Uncovering At about nine hours into the event, the core coolant level again decreased, possiblity down to 7 to 71/2 f t. from the top of the active fuel level.*

The core remained uncovered at this level for about ene to three hours, after which the coolant level was again raised and covered the core.

The low system pressure (s450 psi minimum), the rather lengthly period of '

uncovering, and the additional length of fuel surface uncovered, undoubtedly resulted in additional fuel system damage due to Zircaloy oxidation and embrittlement (followed again by more fragmentation due to thermal shock during the recovering of coolant level), although the amount of additional damage is presently unquantifiable.

Fuel System Damage Summary The picture of the core that has emerged is that the core configuration currently consists of a basket-like shape of relatively intact assemblies that surround a central region of severely exidized, and probably fragmented, fuel rods in the upper central part.

The fuel f3G - E?4G

• Sased on inioimation received via telecomunication from B&W ( April 3)

. rods are less damaged in the lower central part of the core.

Al though the fuel rods in the upper central region may be completely fragmented, the guide tubes, grids, and end plates are believed to be intact thus providing a skeletal structure which supports the remaining portions of the damaged assemblies.

.'artial flow blockage caused by accumulation of fuel debris is thought to be responsible for continuing elevated thermocouple readings. The assymetry of the incore thermocouple readings suggests that a region of the core is more heavily damaged than the average.

O e

O 80-247

gue 4

.; uj u D. W. Labelle

,s.

.f PRELIMINARY SEQUENCE c;..

0F EYD(TS.

(TMI-2,3/28/79 INCID,ENT).

The folicwing sequence of events for the THI-2 incident of 3/28/79 has been formulated by B&W engineers using available plant data. This chronology has been constructed from numerous sources and has not been totally

, confirmed.

It cay n)t be precise in either event occurrence or sequence.

Erent Tire, Ninutes Prior to The initiating events could have core from numerous nostulated turizine trip causes.

For purposes of this sequence, they are relatively enimpotnt. The prime effect is that it led to a los,s of cain feedwater (MFW) booster purps.

O Main feedwater pu.ps are tripped. Alcost sinultaneously, the turbine. trip occurs.

' 0.10 Pressurizer pressure increases to the EF0Y setpoint of 2270 psig.

0.15 Secondary side pressure peaks at 1070 esig end is lici,ted by steam relief valves.

• ^',

0.20 RC pressure trip setpoint reached (2355 psig at hot leg ~

tap) and system pressare peaks at about this valu.e.

Indications from pu: p discharge pressure are that. auxiliary'

' feedwater pt :ps (oce turbine driven, tuo electric) are running at this point; however, no level change occurs in steam generators.

0.25 Pressuriser level peaks at 255.'in'ches (indicated) and starts '.

to decrease with system contraction.

0.30 Quench tank pressure is increasing.

0.90 Pressurf aer level is at a minier.m of 158 inches and~ starts to increase.

Hot leg temperature is at a minicum of 577aF and starts to increase sicwly.

1.0 OTSG 1evel indicaticn en the startup range is 10 inches.

OTSE pressure holds at.about 1025 psig.

2.0

~ OTSG pressure starts a steady decrease. HPI ficw is initiated by ESFAS on icw RC pressure (HPI setpoint = 1500 psig).

3.0 The quenc5 tank's ' increasing pressure levels off at 120 psig.

Relief ulve setpoint is 150 psig.

g-g

~4.75 The hot and cold leg tegeratures start increasing at a more rapid rate.

Analytical simulation indicates that this occurs trhen the HPI is turned off. Site infor=ation notes that operator terminates HPI full at 5.1 minutes.

, - r E-t q-j Event Tire, Minutes _

Pressurizer level indicates a slowing and then continues to 5.0 increase as the het leg ter:perature if increasing.

Pressurizer level l'ndicates a' full pressurizer and the 6.0 quench tank pressure increases beyond the relief valve setpoint if 150 psig.

b RC pressure rodhes a minimum of 1350 psig with a hot leg

~

tecpreature of 5840F. This indicates hot leg is in satu, ration condition.

j This Auxiliary feedwater ficw is initiated to both OTSG's.

i is indicated by icmediate OTSG repressurization to $1025 psia 8.0 and OTSG level change.

RC pressure peaks out at 1500 psig and starts to decrease.

9.0 Hot leg temperature peaks out at 597aF.

Pressurizer level indiciation is TOstored.

It stabilizes 11.0 out at 375 inches at 15 minutes.

Quench tank pressure drtps suddenTy, indicating the rupture 16.0 disk has biwn (setpoint.= 200 i,25 psig).

18.0'

.The decreasing RC pressure stabilizes at 1l15 psig.

22.05 The RCS temperature ' stabilizes at a hot leg of 553 F and a cold leg of 5480F. The terperature decrease from start of i

auxiliary feehater to this stabGization represents a 2000F/hr cooldown. Reactor builoing pressure is 1.4 psig and increasing.

Two feet level is restored in both JTSG's.

The startup level indication shws OTSG B level increasing 50.0 and OTSG A level decreasing. Pressure increases in both OTSG's.

During the 22-50 minute period, the system paremeters have 60.0 stabilized in the saturation condition of a pressure of $1015 psig, temperature of s5500F. RC flow indication is decreasing firm 60 (initial) to 50 x 105 lb/hr. The reactor building pressure is 2.2 psig and increasing.

73.0

' Two RC pumps are tripped (in loop B).

Reactor coolant flow rate decreases in Loop B.

OTSG B pressure drops from 950 psig to 140 psig in 18 minutes.

78.0 Races de core Ws about W.

S0.0 T

folles T hat sat.

100.0

.Both remining RC pu=ps are tripped.

in less can 114.0-120.0 T

and T m em e rap W.

Thot >

hot cold 15 minutes.

8(J - g43 Site inforr.stion notes that ENV relief line was isolated 132.0 initially.

RS pressure starts decreasing more rapidly.

0.,

• ~.f.'. '

?..

Event Tis:e, Minutes RCS has depressurized to 670 psig and RCS hot leg 135.0 0

At 620 F, system temperature is at caxicum scale of 620 F.

would have superheating at upper elevations as long as pressure uss below saturation pmssure of 1772 psig.

RCS shers rapid re-pr.essurization.

_. OTSG B level rarped up from 5% to 65': in 48 minutes.

150.0 OTSG B main steabisolation valves cad turbine bypass

[_ _ _

180.0 valves are closed.

RCS pressure peaks at 2r20 psig.

180.0-204.0 Regulatfon by EMOV bicek valve reduces RCS pressure.

IGI cores on (1600 psig signal).

~

204.0 216.0m HPI pu=2 le to Loop A turned off. RC pressure decreases

. stepwise. RB pressum increases steprise.

290.0:n RS pressure hits 4 psig.

Building fan cocier comes on.

(4.83 hr)

RCS pressure increases rapidly frca 1250 to 2120 psig in The EMOV block valve is closed, one HPI (1A) 318.0m (5.3hr) 35 minutes.

- is on.

OTSG A level is racped up from 50% to 95% on operating range 354.0 (5.9hr) in 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> and to 100% in 1.5 hour5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br />. OT5g A pressure starts to decrease toward zero.

The EMDY block valve is cpened.

RCS pressure starts to 450.0 decrease (2050 psig to 480 psig in 1 hr, 45 min).

(7.5hr)

RC system pressure reaches 600 psig ' core flood tank 519.0 (8.65hr) setpoint.

SSC 9 RS pressure spike to 28 psig occurs.

(9.8nt)

W A reappears on scale, decreases to 525 F in 1/2 hr.

630.0 Thot (10.5 hr)

Loop A incmases in about 5 minutes fmra 190 F to 400' 678.0 Tcold (11.3hr) 750.0 HPI flow increased to 400 gps.

T in Loop A decreases.

gg (12.5 hr) 810.0 T

Wp A kreases.

cold (13.5 hr) 948.0 Pu:@ IA is started.80-250 (15.8hr)'

O e

m - ~ --

- - = ----^

par e_ -

a-

~

. :. 8, e

~-

' e'.

),

4 r

e Tim, Minutes _

Condenser vacuta re-established.

Thercafter SG-A begins steaming to condenser.

RCS cooled to approxi,ately 3000F,1000 psi.

T Letdown line ceased to permit flow and relief valve being used (estirated 14-16 gpm flow).

0 6

Some fuel 'ncore titermocouples reading about 600 F.

RS pressure belcw 1 psi.

High radiation in reactor containment and auxiliary building.

t 4

G e

a e

e e

S e

e e

e e

e G

G P

O e

e e

e E

e e

e e

4 6

o*

e G

e e

G e

80 251 e

G 9

\\h h

n s

9 4

_ h

j.

• s

's %

4I f}

QN g \\ <~,

l k *'

-t (0 s

- 'j

.s O

1 j

m-3

%b

- %s

/

1 J

t,.

(

?

s

~

4 s

9 x

s N

b i'

o h

N y,

K

^

%q

~

.)

-- (,..

C* N

< *~

_a A

y

[

g g

.k 4

g a,

(

wh

.JW A

3 j,/. >

wI

~

4

)

3 8

't. >

1y cC:c:.

y v3

)3 9

s

.~

. )

m%

...n.

- s

~

Q ed 2

~

%g 4

~

se-252

.i

\\

\\

1 1.. -

~ c..'

t O

t o' ~ ~ ~

~~h o _

o n

o O

O

.o o

O M.,

%"N O

o c,

9

f (l} ' v 5 4

b)> p M ?

1* E Core Coolant Conditions At 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> after turbine trip the core had become o

partly uncovered and remained uncovered for about one hour.

During this period activity alarms came on indicating o

significant fuel failure.

~

o Core was ecovered when high pressure injection pump came on.

Two additional periods of extensive core uncovering folicwed o

at abcut 5 and again at 9 through 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> after turbine trip.

9 N

80~253

.. ~

9 tlumber of Fuel Rods uith Defects Based on measured ccolant activity, all of the fuel o

rods probably released fission products, Amount of hydrogen released from oxidation of cladding o

(metal / water reactions) also indicates all fuel rods are damaged.

~~

Maximum Fuel Temperatures Calculations based on Fission product analysis indicate fuel o

temperatures of 1400 to 1600*C.

Heat transfer calculatic.ns indicat'e temperature of about 1100 C.

a 9

The melting point of UO2 fuel is 2840 C so that core meltdown was not approached, The absence of Sr and Ea activity.in the coolant confirm the o

avoidance of fuel melting.

9

~

W 80-254

Extent of Fuel Damace

~

+Y Hydrogen balance calculations indicate from 15 to )d% of the Zircaloy c

cladding has been oxidized.

Continued operation of_incore instruments indicates that fuel e

assembly structural members remain intact.

Absence of silver in coolant suggests that control rod materials 3

remain in place.

Cont. nued low thermccouple readings at periphery suggest that i

o peripheral fuel assemblies retained much of their original geam.

The picture that emerges is that the upper central part of the core e

is severly oxidized; probably fragmented, and largely confined to the core region (based cn loose parts monitoring data).

Partial flow blockage caused by accumulation of fuel debris has o

probably occurred and is responsible for elevated thermocouple readings.

s80-255

9S2 OS

/.c f

W

\\

,f h

/,~

.s

,h J.; y

,,/ t

~

b

's gs s.

.. g' p>

s

~

>l $@f g

h Wf h

S

~

3

(

q

.\\

E

/

&p a

Y p

f c.

s

{

q'.f i

?py f

e5

^ \\'

A} l%,,

~

s n

'/

f p/l(Q.

g i

lb Y<

n g

f*9 a

(

3

+

r E'

• f&

~.

c}"* t,4,'g-y, y,

4

/

e

,\\

,y~

,w 6

g~

2 g) y e

d$

~~

R

.1, ',

t a

.?

d ap;,

s yLs"e.,

np

~

. V/

s p~4,,

1. ~

.-I 1

?

s g

S o

a.

s: co s:.

E' l

7777777

l':

i lykD l

/

'N

/,

i

/

/

if

~-

I 1

=

f}

wx ur 9 -

y yr--

s y

, li iO O C)0 00 de3 h

e O( )O i

<< m r 4

. ~ - - e" ' '-

b E

l t

'.a m cr

-<nz T y%

,_ -tr

- ~4_

o=u_ne; m

- n an

, y b',... \\1

.x

.,i

(,~=,~"

I T..

.7 M 3 FUEL DEBRIS 1 'K N

~

' x,-1 g -uw,,

'I 1

a; t,

1

' r asuusa.

g l

b S lN \\J s

li OUTER ASSEMBLIES-s

]y$ }'

)

1 t

T y g'

~

k /

r FUEL CLADDIllG PERFORATED BUT j

i N,s*, ' us m a =

RODS IliTACT

n N b

b

\\

l/-

f[

n f jf 1 su sY s 'n f

sc.y,,mm 1

E ALL GUIDE TUBES,

f AtiD GRIDS It1 TACT-d 3

r

M jf

:j

fY jou e-sm u 4 / d h h kI/ 5-1 t 'f ' ' Idl,2-#174 Q. ~ ( .r 'bduTM%$hhfrW f' ~ u i ~ Nj h I ,,. y-c..v ues C cf %l I y Q ,, / % )y i \\,- ,:m,.cwm ' s, 3D Q)2 N a x n xs I 80-257 REACTOR L'Eb.El (i:11ERN'l3 4.f.t.C/.tRI / .}}