ML19224B917

Submits 790401 Record of Action in Determining Requirements for Emergency Response Action Assuming Core Melt Accident Would Occur

ML19224B917
Person / Time
Site:	Crane
Issue date:	05/31/1979
From:	Blond R NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES)
To:	NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES)
Shared Package
ML19224B914	List: ML19224B913 ML19224B915 ML19224B916 ML19224B917 ML19224B918 ML19224B919 ML19224B920 ML19224B921 ML19224B922 ML19224B923 ML19224B924 ML19224B925
References
TASK-TF, TASK-TMM NUDOCS 7906280007
Download: ML19224B917 (13)

Text

.

%o, UNITED STATES NUCLEAR REGULATORY COMMisslON

~

WASHINGTON, D. C. 20555 k..... /

May 31, 1979 MEMORANDUM TO FILE FROM:

R. M. Blond, PAS, RES

SUBJECT:

RECORD OF ACTIONS, THREE MILE ISLAND INCIDENT On April 1.1979, the PAS staff was called upon to determine the re-quirements for emergency response action in the vacinity of TMI assuming a core melt accident would occur.

Efforts were initiated to determine the following: the sequences of events which could lead to core melt; the timing associated with the sequences; the magnitudes of radioactive materials which could be released; and the dose versus distance rela-tionships that would be associated with such releases. With this information, contingency plans were developed which w!re designed to protect the public health and safety in the event of a serious accident.

Appendix 1 dcscribes the major failure sequences and the avent trees that were developed. The timings that were associated with the events are also given in this appendix. This work was done by Joe Murphy and Mark Cunningham with consultation mainly with Rich Denning at Battelle.

In addition, a short discussion is given on the possible core melt / concrete penetration for TMI that was done by Rich Denning.

Appendix 2 discusses the assumptions and techniques that were used for the dose versus distance analyses.

Roger Blond and Peter McGrath were responsible for this work.

Appendix 3 describes the sensitivity study that was done to show the impact of the proposed evacuation strategies on the consequences of a severe release at TMI.

Roger Blond was responsible for this work.

Appendix 4 describes an attempt that was made to estimate the released source tenn for TMI. This work was done for Harold "eterson by Roger Blond.

7906280 w y gg7 0 6 l

. Appendix 5 presents a study that was done by Roger Blond to show the uffects of time af ter shut down on the possible consequences of the worst credible accident at TMI.

6 NY

~

Roger M. Blond i

Probabilistic Analysis Staff Office of Nuclear Regulatory Research l

Attachments:

t 1.

Appendix 1 2.

Appendix 2 3.

Appendix 3 4.

Appendix 4 5.

Appendix 5 257 097

APPENDIX 1 MAJOR SEQUENCES OF EVENTS mit

w Major sequences evaluated here are tied to the loss of forced circu-lation in the RCS.

The ss of flow from the reactor coolant pump (RCP) is the generalized initiating event from which other initiating uch as loss of offsite power can develop.

j APPENDIX 1.a SEQUENCES OF POSSIBLE SYSTEMS FAILURES 1

1 J

Figure 1.b-1 shows the loss of RCP event tree.

This tree shows the various options available given the loss of the RCP, and indicates which combinations of events or failures would lead to core neltdown (CM).

The sequences denoted with an asterisk are those which would be ex-y pected to follow the core meltdown progression discussed below, leading to the variety of atmospheric radioactive releases and consequences I

discussed later.

Some core meltdowns could be expected to be delayed for roughly a week because of the availability of ECC injection over that period.

This method of core cooling, hov.ever, is not expected to a

be adequate to prevent cor,e melt; e i such a core meltdown is assessed to occur at roughly a week.

A rough measure of relative probabilities of the various outcomes is indicated by the notation of L, M, H (low, medium,high)'.

The column on the right-hand side of the page indicates the relative probabilities of the sequences, with "LM" as the highest 3

probability and L M as the lowest.

257 098

m r-5-

,3

. __. y,

_ -...,g.. _.7,_ _5.__ c.._.,r,2.

y

,,)

g

.__g j

j

-- _. - y.... \\p,*__

y.

.e*

..N.._..........M._.c__......._._

,M

,......__....n_._

M..... _.

g_.__.3.... __..____ _.._.

3 4- _... _.

.=

. _.. g. _ u

_ g.... t...

4*

,t

... y.. _ _. V

y v

g 3

6

. g.) _ f _..

.g ;.

[

g g

o.

o.. g o. o g. _ g. o..a.

e, g

1

/

.a t

{

s

3...

e t.

1 i

L 8

.J i

a p

1

v.,

./

d b.i.I J

J7 Ju r

q 5

c.

g v

g, e-

\\,

r g

.p c

~. _

<.'. (

f

f g

5 s

h

_._I o

J A

4 h' e\\

0 A

J a

d

. g j

, 5

'r.!-

.d o - t.

5

.A. $ ?Y M

e k, 1 257 099 I

L

.c

~

N APPENDIX 1.b MAJOR EVEtlTS AND TIlilf1G IN EVENT OF CORE f1ELTDOWN

.p ans W

Event 1

- Sprays and Coolers Operative m

Time =0 Flow stops, core and water start heat-up Time =100 min Cor'e starts to uncover

.]

Time =150 min Core begins to. melt Time =200 min Molten core is in lower head of reactor vessel, pressure is 2500 psia Time =210 min Reactor vessel fails, containment pressure goes to 25 psia Time =210 min Hydrogen burns, containment pressure goes to 67 psia Steam explosion possibility - minor consequence CONTAINMENT SURVIVES (Failure assumed 130 psia)

[

Time =10. hours Molten core has melted about 1 meter into basemat Time = days Major problem - handle hydrogen, oxygen - maintain contain-ment integrity CAUTION:

- Keep sprays running e

- Keep water many feet over molten debris j

- WITHOUT RECOMBINERS Hydrogen continues to build up BASEtiAT SURVIVES Event 1

Conclusion:

This event should not produce major releases Event 2

- Sprays and Coolers Failed Before Flow Stops S.'

Time =0 to Time =210 min.

Same as Event 1 - containment pressure is 25 psia m

Time =810 min Containment pressure is 70 psia Time =1 day Containment fails due to steam (mostly) overpressure -

about 135 psia C0ttTAINMENT FAILS gg Event 2

Conclusion:

This event leads to major releases.

l 257 100 l

i0t 192 1

seen O

The event tree for core melt leading to various releases is shown i.1 Figure 1.b.

I The following are essential.in the event of core melt.

l.

Sprays and coolers are required to prevent major releases.

2.

Hydrogen must be recombined or otherwise removed from containment.

I 1

4 hrt i

I A

8k,# '

1 t~

l

..p;;.

T

?

dUl L6.

a i

-~

..... ~.

I i

m

..y g

h

._@_.6..

=. *....._

'8_

6

.a.._.h m

,,w, y m,,,e y w

..... _ _...e

.-..s

'i

.e

.s

.g MN_h_

____g

-E

--N_

h__'

• O

""8'E"*

O*

N_

e.

g yg.me-_6_..

e W._

_W_hh.h.

__ _.. =

6_ _......

._ga 4..-.. _. _ _ _

F___....___.___

_ __ _..... I _._ _. _. 5.

SJ A, _

.,/.

__..y

.. f." _ _..._.. f,,.

J Y.-..

s.

.m.

..._..-g-

.W

)

y.

./

_.... _s.

__ _.._ /.._...../*

9 4

6

.1

_. _., q,

.y

.y.

v*

p.

y a

y y

g

_._.g...__.......

p M

g g

.j

._o...eu.

e p _ -

g e

...I g,1 1

o t

y.

CC.

g C"

t

..p.

i O.

l __. _..

i,', ;. -...

l.

c

. c;,.

//._

s

._4

.....9_

4:....

.m

._.gs..

._.g.

...m

_ w.e - w w.m -.....

j. __...

_.___.__._._./__

_.N h*

-M..h.*

m.

._......_...*.._.._...___.-.__-.ee.6

........ - -_... _...]

C N#

.._...... 4..

..o b

1

(. g Ci

!y pg.,.g

..._.._m 6..

ggD b.Wh.lS '

A

~

rn\\

lGl b

h)

<1 emas Ic.

Large Leak in Auxiliary Building (AB)

The activity level in the reactor coolant is so high that substantial

~

releases can come from small amounts spilled in the AB which requires

".)

once through ventilation.

A leak of 5 gpm to the AB atmosphere is assumed for the expected level of lea'kage.

A leak of 50 gpm is taken as a large leak to consider a major leak in pump shaft e

sealing or some similar mishap.

Based on the leakage experienced already only the noble gases and no iodine are assumed to evolve.

4 The AB ventilation exhaust is assumed to flow through the charcoal filters.

{

Id.

Hydrogen Exolosion in Reactor Pressure Vessel A detonation of thu hydrogen oxygen bubble in the reactor vessel 9

could rupture the vessel and/or crush the core.

Rough analysis indicates that the pressure vessel would not rupture.

Postulation 1

J of the core response is difficult.

If the core is crushed, it could I

effectively prevent core cooling leading directly to the core melt 4

sequence described earlier.

It is unlikely that compression would lead to criticality.

l

't i

I q

i

701

/S2 CONCRETE / CORE MELT PENETRATION From the modelling of concrete penetration in MARQi, it is not possible to conclusively state whether or not the molten core will penetrate the basemat of the containment. The INER routine of MARDI is primarily applicable to the initial period of concrete attack and is not reliable for predicting long-tem behavior. Tables 1 and 2 illustrate the predicted penetration as a function of time for two of the MAROi cases.

In addition to the difference in decay heat for'the two tables, the core was asstrned to be quenched (to about 2500 F) prior to concrete attack for 0

the 14 day case whereas the core was at approximately 4130 F at the beginning of concrete attack for the 3-day case.

TABI.E 1 CONCRETE FENETRATION FOR ACCIDENTIgITIATEDAT TERzu DAYS i

T. hr E. ca R. en 0

0 117 2

3.1 204.4 4

37.9 230.1 6

36.2 244.4 8

136.1 253.8 10 183.9 260.2 12 185.1 263.6 14 166.6 268.6 16 187.8 273.7 18 188.s 278.8 20 1st.S ss4.0

• Tewperature of core at begin-nina of concrete attack was

- 41300 F.

l TAnLg 3 CONCRETE FEMETRATION FOR ACCIDENT INITIATED AT FOURTEEN DAT5*

T. hr

1. em R. ca O

O 179 25 33.3 237 51 48.3 278 8

100 63.7 343 200 79.4 4C 3

250 84.0 487

__-=:e===========

Core van nosumed to be quenchad to - 25000 F prior to concrete attack.

- n..

Appendix 2 Consequence Calculations (Dose versus Distance Relationships),

As the event trees in Appendix 1 show, four RSS cora melt categories were identified as being credible given the status of THI-2 on Sunday morning, April 1. These included the PWR release categories 4 through 7.

With these release categories the basis for the dose calculations, we considered the following specific data for THI:

the inventory available to be released, and the sequences of meteorological conditions that were being forecast for TMI over the next few days.

Core inventories were obtained from an Origen computer run which approxi-mated the status of the TMI core as of April 1.

That is, a PWR core was used with approximately three months of operation and about four days natural decay since shutdown. This core condition is equivalent to less than one fifth of the core inventory used in tha Reactor Safety Study.

Meteorological conditions for the site were developed from discussions with NRC meteorologists, and local meteorologists at the site.

Two sets of meteorological conditions were found to be likely for the Harrisburg area over the week of April 1.

The first meteorological condition predicted light winds; stable dispersion conditions (3-5 miles per hour, E stability) which would be expected to occur at night. The second meteorological condition predicted light winds, neutral dispersion conditions and rain (3-5 miles per hour, D stability) which would be expected to occur during most of the days.

These two conditions were assumed to be representative of the week of April 1 with the only exception being infrequent locally heavy thunderstorms with associated high winds.

In order to have some assurance that the doses whicn were going to be computed could be used to make judgements for emergency response planning, we assumed no special protective measures would be taken such as evacuation or sheltering. The doses reflect the accumulated exposure to an average adult for a period of seven days assuming normal activities.

All other assumptions used for t'hese calculations are described in "The Reactor Safety Study, Appendix VI, (WASH-1400)" and " Overview of the Reactor Safety Study Consequence Model (NUREG-0340)." The doses that were calculated by these models and assumptions should be overestimates of the doses that would actually occur, but the techniques tend to be more realistic (smaller doses) than licensing calculational techniques.

257 105

a Accident Effects

+

The attached table gives the expected doses to the thyroid and bone marrow for 1/2, 5 and 10 miles from the reactor.

The bone marrow dose is dominated by the ground shine exposure pathway over the seven day period, whereas the thyroid dose is dominated by the inhalation exposure pathway from the initial passage of the radioactive material.

e 4

e t

7 i O a

Dose ** tem) 1hyroid (4 Jay) hhole Body Release Miles 1/4 5

10 1/4 5

10

'V II

• 10

.05

.01

.2

.002

.0001 llox 7 15 1

.1 5

.1

.04 Box 6 35 2

1 10

.3

.1 Itox 5 150 10 3

20 10

.3 Box 4 500 25 10 50 3

1.5 Box 3 2000 800 350 70 30 12 lHi* Containment not violated.

Continuous release at design leakage (.11/ Day)

(See Joe Murphy)

• • Dose in rem with following asstenptions:

1.

7 day ground dose exposure 2.

Actual meteorology at time of accident 3.

Actual inventory at time of accident 4.

All other RSS assumptions 257 107

Appendix 3 Evacuation Model Sensitivity Study Y5 To determine the effectiveness of the various evacuation strategies that had h

been proposed for TMI, a sensitivity study was performed using the CRAC model Five for the largest credible accident and the TMI 1980 population.

" precautionary" evacuation strategies were evaluated to be able to compare the total whole body person-rem (within 50 miles), the total latent cancer

~

fatalities, and the evacuation costs that would be expected from each.

If the population was not evacuated they were given seven days ground dose exposure with normal assumptions for shielding. included no ev The five s, 5 miles, 10 miles and 20 miles of the reactor.

Total Evacuation Evacuation within x Miles Whole Body Person-Rem Latent Cancer Fatalities Costs #

0 1.21 x10 93 0

6 5

6 89 2.68xlg 2

1.18x106 79 2.9x10 5

1.05x10 7

10 7.45x105 56 1.65x10 7

5 19 6.42x10 20 2.58x10 There would be no early fatalities expected from the TMI-2 accident under any of the evaucation strategies.

257 108

~

Y/ay n

~

b. z J i k h rz- +

5..

h r I -- ',

Appendix 5 SENSITIVITY SnJDY ON TIhE OF CORE MELT AFTER REACIDR SIIH DOWN FOR Dil-The enclosed table gives a sensitivity of the censequences of a severe core melt accident as a function of time after shut down of the reactor. Calcula-tions of dose versus distance, man-rem, latent cancer fatalities and early fatalities were made for five times after shut down. The same assumptions made for the calculations in Appendix 1 were made for this study.

~

To better understand the effects of the status of core bum up on the results, the last row of the table (RSS at t=0) is given. This result gives a comparison of the RSS asstmtion of a core near its end.of life versus the DfI core with three mont u of operation. There is approximately a factor of 5 reduction in the long lived isotopes (Cs, b) between the RSS core and the core at DII.

g.eh.

g_

m

'h*

g

.a.M m

hp *

=e.-

t i

.+J

.=

t s.

..A

.u...

t

.*e.-

L. l... - _

-...-2. l~._

~

Y--

~

l..T..

-]- -

2..~- - - - j.

- - -- ' % <:;f

~

^

. 3...

257"iO5~~

~

~

~

~

l L.

-. L-

9 MlOLE BODY 11rfROID M-R EARLY IATENT TIME 1/4 5

10 20 1/4 5

10 20 Initial (t=0) 1930 174 48 9

31520 1541 414 74 1.4;>6 177 114 t=1 Day 614 103 28 5

15860 827 225 42 1.09+6 68 87 t=4 Days 293 63 17 3

6752 374 104 20 9.48+5 8

77 t=14 Days 297 23 6.5 1.2 2363 133 37 7

7.3+5 0

58 t=21 Days 56 14 4

.75 1272 73 20 4

6.7+5 0

5.3 RSS at t=0 2250 113 60 11 36830 1715 490 87 2.16+6 207 178 e

w N

M 1

A

~

.ps ang

/'

q*%

UNITEo ST A TES 8

i '.

NUCLE AR REGULATORY COMMISSION

{ ;% @D)i WASWNGTON. D C 20555 I

%..'.f /

I April 24, 1979 L,__

\\fffbf MEMORANDUM FOR:

Harold Peterson, Jr., RHSB, 50 FROM:

Roger Blond, PAS, RES

SUBJECT:

INITIAL ESTIMATE OF REl. EASED SOURCE TERM FOR TMI The enclosed figures give a very rough estimate of an initial source term that was potentially available in the auxiliary building for release to atmosphere.

The following assumptions were used,to make the estimates:

1.

Sump pumps in containment could potentia 11.y pump at 200 qcm or a total of 12,000 gallons of primary coolant were pumped per i

hour.

The curves, however, are for 8,000 callons of primary coolant.

Points are also given on the hourly figure assuminq 40,000 gallons of primar.y coolant were pumped at 8,000 gallons r

per hour for five hours.

2.

Inventories were calculated as follows:

The Bettis Analysis of the primary coolant (including makeup water)showed6500uti/ccofI-133t248hoursafterthe accident.

This is equivalent to 31,450 uci/cc at the time of the accident; assuming no dillution*.

From a ratio of Origin core inventories, we estimated the I-135 concentration in the r

primary coolant to be 29354 uci/cc at the time of the accident.

This is ecuivalent to 830 ci/qal of I-33 and 770 ci/qal of I-135 6For 8,000 gallons of primary coolant, there should be 6 curies of I-135.

The curies of I-133 and 6.16x10 6.64x10 Xenon source was assumed to come solely from decay of the a

iodine.

I I

Roaer Blond 257 111 l

ourticart oocuxenr

. Makeup ' water could have dilluted t Entire document previously entered into system under:

H &Tllool3/

^"o 7TOs/C s

no. of pages:

=

i