ML19290B982

IAEA Technical Assistance in Nuclear Reactor Safety: Philippines, Prepared by Philippine AEC

ML19290B982
Person / Time
Site:	05000574
Issue date:	07/31/1979
From:	Willis C PHILIPPINES
To:
Shared Package
ML19290B974	List: ML19290B973 ML19290B982
References
TA-PHI-9D-05, TA-PHI-9D-5, NUDOCS 8001090009
Download: ML19290B982 (47)

Text

--

e ; *.. *

.,..(*....

d

... ;,f,q:P -

y..

• ge

...., i. *, g

,.~..

rn S"**

e'

.' %;,.*, 4 4

-.. :s,.. k., n*...,*'.,'.. * % s...' '. %sk.% '

' (. A.'-

n.. '.>q

&'e-

.-.4..... v(.y.

-:a

?T-

.v.

.~,.. :. /*,

~a-~.e o

1

... ~..

t

.4

% 1 A.

.%., g.

o.

5 *.

,....,e+-.:.f

.m

.m..r :.v'%s ".,.., g,:.,3n. m. ~-

- w -:

...,a-

- v -u.

t m:

r~--

,,s.,

s..

%. W

.g.,w.e....; @.-

M. 4: g' i: e,d,:%.,... -.

.t

.s e

?

s..

s, 2

..xxp

..... -. Oc....

, e ;

,t<. s... e.O y..-

m '.

_.... ~.,

-p

. ;.. e.,., i,. A y. e. c -.., '. e. - ~..g,. :. 9

-,- w m..

N.. # - * *.

,l.d.

.g a

, a s

. :., v Q.~.

• e'*:

.wg,,-*

5

,..,e

.e u...

.s,.

>, - ~

. e....... v ;*. s :. ~,.. m-c...,

~. <

n.~

~

.;3,

..e..., s.

..a.

c,

,. t

, n

. m.

v.

a, 1.

z.

,p,-.,..

... ~.,.

.c -

r y

.-&..~., g,

,~

/A-.f. w.IAEA TECENICAE ASSISTANCE IN j

.,w.,.,.

... c u..

g.

.m.,.,.-...:

.4;,a...4i.NUCLZAR. REACTOR SAFETD PHILIEPINES'-

)@c.

~ ' - '.

i

~- k

- mW.W.n r,....g.

n

_.,, -. ;s +.

.-..+,._..-..~,'sm....r.g,..,.

.s.

,5 a,.,.,..

v.

.~

ow,

e..

u,-

e o u..

.,.2..

-.,%'..'.'?.

..* % *. *; p 1, b n.: w 4.*

.' *i'.*e 1.,. 'i,,.'e.

. =

.6-e.--4

• *wt,

+

-.-i<- '

t

. c*

.f. ^

's..*,..

a.

.'. -...k~.*i..-p~.'..N.,,

.* -.., --g,1-"*

r s-

.u t

'.a ***.o-g.'.4.?."', %,,.,s

.3.,...e.-

,4

,.v.6

- ] gj. 2'.a.~.6 7.. ~.

'..y

.s-

.f.

^;.

• L. ** 4,.

t. ' *.. !.V

.*.y.j.

4.;. s,.

f 7

tw

,. T '%..m*'.;.,. y: )'

h. :a t. ' /.*,.'.,.

• r.,-

~..

.-...i s ;. %..< pi.x,,,,.

ir

..so 2

....,.. f.. m :

,.,e. *4

+

,, g, r.

. e. J -.,...Y s

y ~.o,e -

e

.A..y..a iy9..

, +

,e

~.s.+..~.s s

1-* g.;; 3., r *.7 s

a

.' g e.

.*==.- -.,,.

m

.1,?*

. - s *. q.%..

. i,w.. &.....

• ' s. ' J. 3 _ _ ~,,.:

• =

r

,,n..

g e..,..,' n

.,\\*

.e s..*

,.e i.

t y

p..

_F

.- a w, J.

1..,*.

..> y * ] b y ' '.-'.e..,. <..

.u..

..a re

., ~

o w.

. ~ " :s w. p. '. -

, Q Q.,-....

.. w. - -

- ~

z, M y,..:,.

..m

.n.~.

G - u. -

~

.w 7,,;r,

..'w

....,s.5..- n/ PHI /9D/05&

,.,. d..'f I

m

-.=.

+. -

..r.c e -., - y' ; :.y e..

,"~

,s

. -. -.ma - g 2.. ';,s, @e

,,>>.G*..

,. April > June!.1979.(,..

• e

^e s

.s w

.o w...,

s'

, * ' $...-.., y

,,.", t..*%.ya ' e ; '. * -

'- {. * (. _

4 e.

S' r.

wc

.y

_t.'

n 4~

-e*,w,,.

e

. J. e g *-.,. a e..* *3., m'

~.~.

.e s.,

o s

4 t 45 -*,

...-~s--

j,~<.

g _., e g.

M c

c. -.; e

..?., ',s..-

~

'.H 5

_,.e,-

• 6*f4 ;

o y es*

.o

p..;-

4.,. -

.f.

.,v.

'r.

a.

^h}..,,,., q y - y

• M.e*g,.

.., $,e.

.g / -.,

s s-

,,.L e -

.a.t#

.n,.i.,

.,.s....q-e n... g....e. =ym

.6-

..s-

..r-

-... v.

., 9, ;...

.....v

~ -

.- s e..

m.s r

.. 2

..~.3

-,. Q.,1..,.,2...,. :,. -

a t

O

, 9. h.r..,

'nm. A.9'II,I IS : ' ?,_'N'.

f yt..

F.

W 3,. e..;.c.

t.s.

-L.'

l

..g.W.

e aar,Reacter Safety Experr c

. m,

.,e

. - * *-. '.. J.: r; ;,i.,. -

~

Yr.

s' y.

.s.,

s' m

'E,.s,pp.

g.*.

.l..

-z x.

.g

/

-z.<.s7. y,

.3.

.~ 3, a_n..,...

p %g

.n

.c a

.y,

r

+:; -

.......,...'.,v.-.

.r.

..= -;

...7....,*,.._g,.

s.

.c

,(

.,n.,,.,,

. y

.~

. :.c. -. u, s. -..

s

.n

.. 3,;.

..w.

. -- f.

r,e:.

,O. ;.

y_

<. :.......,s : v y,,, x

. <<.e.m.

a. v,...

.1..-

.. y.. r..

'x p.. :s.,3.,.. ;,., y.,.

s c

~

. z.. -

~

..cm q.,,'; **'.,e,e' ~'%

s4 4 ~ *!f,,, * :

s

.a;

.e-a

. t, s..

<n

.>.,-4,.,>...

~.4

..,....a..,,,...... :

s e..

.;,.c-s

.r

1.

y*.y w a._~...,

., g.....

.v-r...,...

.t

...,.s..

...... J s, a

n.

r s.

c.

n.

+-

m-

^ '

.s

~. -+. !W,.A... d;. -l *.

.y m,

- %n.e.:,; ~..., u

'. b ' :.

N.

~

c.'...,

..,.3...

c

• s.

1..

.. - c. m.

.pr n.,,.

l

~

, l.. ~.,'M,.,..;.'. ht.n..,..,. e Qu.1 f..L9 7 9 ',..

s

., y,..

4,

,. ' i;..,E?W7 PINE AICMIC -ENERGY CCfiISSICN

.a *T.-.

~3

~

W

~

r.

t.

ar

..[ g, '*

s.

P g

r, G.-'

4.,- t',,'Quercu City-s#

.c.

2

s...; -*

4

.s 4.-

..*!. g *.

8 d

p.

e 8 0 0 1,.

f %

V

._.*-g 1729 018

ll (yD 'd A

'D " ~D

%6 W

CONTENTS Overview

. P -3cipal Problems

!!!. Progress and Recommendations During this Period IV, Principal Correspondence, April - June 1979 V.

Accomplishment and Reco==endations in Previous Periods VI.

Working Relationshipc VII.

Fellowship VIII. Equipment IX.

Progras of Work X.

General Consideration and Conditions Appendices A.

Nuclear Incident at Three Mile Island and I:plications for the Philippines.

3.

Reco=endations for PNPP-1 Review Based on the Three Mile Island Accident.

1729 019

IAEA TECHNICAL ASSISTANCE IN NUCLEAR REACTOR SAFETY: PHII.IPPINES I.

OVERVIEW Construction of the Philippine Nuclear Power Plant (PNPP-1) was stopped by crder of President Marcos on 16 June 1979. The work steppage was precipi:ated by concerns about safety resulting frem the Three Mile Island incident in the USA. Public hearings are being held and presumably the decision as to whether to continue the project will be based on the outcoce of the hearings.

Until work was stopped, progress was being =ade on PNPP-1 although the project had suffered a 6-=onth schedule slippage due to the US State Depare-a"t's failure to issue the export license. The work has been of good quality.

The Philippine Atomic Energy Ccesission (PAEC) is a functioning, eff1ctive regulatory agency. A PAEC inspectien team is at the site almost half the ti=e.

The plant owner, the National Power Corporation (NPC) is attempting to be core effective. The turnkey contract severely limits NPC's ability to centrol the project. This forces the regulatory agency to interf ace directly wi:h the contractor, Westinghouse International Projects Corperation (WIPCO), in =any areas.

In general, the prognosis for the Philippine project is good 12 the political problems are solved.

1729 020

. II.

PRINCIPAL PROBLEMS The =ost important problems at this time appear to be:

(a) Philippine public hearings (b) US export license (c) The 1978 IAEA Safety Mission report (d) Effectiveness of the owner (NPC),

(e) Maintain quality through a work stoppage There problems are discussed individually in the following paragraphs.

II.l. Philionine Public Hearines President Marcos reportedly stopped work and called public hearings in respense to a letter frem for er-Senator Lorenzo Ta5ada expressing concsrn about the safety of FNPP-1.

Senator Ta5ada is a precinent = ember of the opposi:ica. His concerns appear to be based en infornation received from Filipinos now liv-ing in the USA.

The hearing panel is chaired by Asaembly=an Ricardo Puno and includes for=er-Justices Ccnrado Vasques and Jose Bautista. Senator Tanada is the principal anti-nuclear interrogator and spokes =an.

PAEC and other Gcvern=ent agencies have made appearancas,pri=arily to questien the WIPCO and NPC representatives.

In general the hearings ar a much like those we had in the US seme yea.s ago. The lack of knowledge on the part of the critics results in hours of pointless questions. Perhaps the =ost significant points to date vera nade by Mr. Wn. Ecuard Arnold, President, Westinghouse Nuclear Intr.rnational:

1729 021

. w (a) The same failures and errors that caused the Three Mile Island incident would cause no damage at PNPP-1.

(b) In any "small break" LOCA the PNPP-1 reactor coolant pumps will continue to function, with no. need to shut them down.

(c) The FNPP-1 shield building is capable of withstanding the impact of a Boeing 707 aircraft.

At this time the outcome and even the direction of the hearings is conjectural.

II.2 US Excort License b'IPC0 applied for an export license in 1977 but the problem beca=e critical only in the last year. By the end of 1978 equipcent was ready to ship. The need for the equip =ent en site has been growing since the first of 1979. Because of the export license holdup, the schedule has been slipped 6 ncnths. If there are further delays, further slippage will result.

No justification has been offered for hol'ing up the license. Clearly d

the Three Mile Island incident is not the prcbles because at least one export license has been issued since that accident occurred. The July 1978 I.EA Siting Missien Report is believed to be an important factor. Other probable factors are (a) the adverse publicity concerning finances and (b) the activitias of 1.e.

anti-Marcos Filipinos in the States.

Cnofficially we hear that the license application has been rejected but there is no indication of what needs to be done.

1729 022

The principal Philippine concerns about the delay are costs and el2ctricity shortage. At (US) $25 per barrel, repiscement fuel for F'!PP-1 costs some 170 million dollars per year (at 70% capacity factor).

Tae major safety problems are those associated with interruption of the work.

II.3. 1AEA 1978, Safety Mission Issues There has been little discernable progress in this area. Conditions are essentially as described in =y March report *.

I discussed the need for further =aterial from WIPCO with Dr. U.H.

Arnold (President, WNI) and received assurances that the material vill be pysvided soon.

I have had no centact with the E3ASCO people since the teeting in Vienna so it is not clear what is being dene.

Centacts have been developed with the CNDP/ UNESCO Regional Seismological Center for Scucheast Asia. The Center has been helpful in providing an independent listing of earthquakes for co=parison purpose. The data are being studidd to see if they can shed new light en the earthquake questions.

II.4. Effectiveness of the Ovner (NPC)

. The owner is less able to control this jcb than normally is the case.

The principal reasons for this lack of control are:

(a) contract provisions giving the contractor (WI?CO) an unusual degree of autoncey,

• C.A. Willis, " Status Summary Report on Resolution of Issues Raised by the 1978 !AEA Safety Mission to the Philippines," Philippine Atomic Energy Commiasion, Queton City, March 1978.

I729 023

. (b) a prolcnged reorganization which kept NPC management personnel unsure of their responsibilities for nearly a year; (c) communication problens within NPC which keep site personnel ignorant of activities of headquarters; (d) general inexperience in nuclear matters (a standard problem).

The contract provisicas giving WIPCO unusual authority are advantageous to the project in some ways'. Most significantly the. quality of the work has been kept high and the work was ahead of schedule until the US export license barrier was encountered. The principal disadvantages have been that,NPC has not had to accept respcusibility and NPC has been denied access to certain docu=ents (procadures, etc.) that are needed for under-standing of the plant. NPC is ccacerned that the contract provisions have limited NPC's ability to control costs.

The NPC reorganizaticr. started in May or June of 1978. A top level NPC organization chart (Figure 1) is new available. I as told that there is a chart for the Nuclear Power Depar.tnent but I have been unable to see a

~

copy despita many atten: pts. Whatever the present state of affairs, unquestionably nany kay people were unsure of their responsibilities for an extended period. Current problema with the delayed response to the IAEA Mission Report and with the US export license =ay be results of the these organizar.icnal problems.

NPC internal ecmeunication problems are evid'enced in many ways.

Examples include the site personnel being unprepared for PAEC inspectors 1729 024

e T

H P

E D

g N

8 o,

CM b

i

, o.

k T

S d.

J 8

AT TSEl -

M s

k R NEC a.

S E

(

tuA ECI l

t E

C k C7

! Ut I -

I I

a 9 DkVl V

J u

9 sot.

Cit 1 ESST t

u WE.

S k! y FR1 n l

t T s Ai A

WA m ENRp 8

OZ u CAE s l

PI n t Ni I.N a

IVtllCC i E r AJ A f Nk uO i

TAN C. T a

H fl i J'

c o

L E n Nt a Td W

H O

Ttn1" D e 1 n NWi r I

r y

l' i

Lt F

Su E c NL a.

tl t

uN3 l

Cf J

r n

kD a AI J ie l N a.

kD

&l l

A !J l

G T k C A

i H

l 1 H" _ TS e t

W f

t i S u S.

u t'

t F

e E H b

Ak rC A L. d'. C.

ll' A C T

N z t T, TuI o

A N A{6V HI LY TI a i.

L k

I' k

AA L N a

a lk NNF luL H

N sa.

I H,'

'b Ck '

t t e J L

p.

1 l

t ID A. C t

N J

t T

b S A luk P h S

l' iEC t

k C t

o Sr.

ut kEi E l L_

d k A CS l' T t t l o

i_

- At

- N T

Lk e H

Lt n_

Coa s

a E

GTu C

o.

"E

T t I l

• c u

4 E C H:

TPJ 1

s Vk!

t V e 1 N t t IV n

,I J g

• _ k P a t

TEl g

I DHn A

V a.

A m

A O i.*

nk a

C S

ln A o

l 1 a

. I bl L

I I t

l kC i

n

• rC h A l

A g

S V

t s

S k K l

t.

t u

uI l.-

k, F.

oJ J

T ct.

i o

S 0 H.

O b t

k c

n u

l A1 C A l 1 c#

6, A L A

N s E

u.

l.

b Al I

l A

T A

TA o

tl u

a V

W tl c

e T

T a

1 u

HV N

U k. I CL lL lpdL t lu_k Z

f I

nliY F _ I S al t s

i t

1, m

. _ A u k

Fbe k

I

. 1 i l'

E i

=

A n

tu) u N.

1 n

l s

un l

)a O.

u i

t.

Nl S

A C a C

a l

L Or i Ni C

?

i li a

Ie mA L IT e

a N

C J

A _

C T

0 L

I d Ua 11 TL n 1

'o l

u I t l

s IC THI i

I 4

4 V

ot o l

a t

T E s l

s lN o.

m.

'l iu C l C

l C O lui A li S

(

bn A

o B

o k

t c

l s

h t (I m

iT a t

N l

hSf l l Y

i AEl k k TkA t A I

C Y lt SA m o

L A

I' A

if t

o.

1 Cu k

N t

k 1

L l'

l'.

t N

t l k.

i ik I.

i H Ll tt 1

3; L l.

l W

C i t A

t bH t

L.

k ;

E se.

O P

(

L St A1k l'

CI T I m Ml Y

V 4'

a h.

V.

t

,11 s__o_sA.

h k 1 C.

^ '

r S0 L

a_ tiuA k Y

ts o.l I 1l.

laWI T.

I S

F 0

A 1.. k(

F.. i s.

r k.

o

'l._ k n.

4tc I

I'.

g_

r I t n _. uTt b s a.

luN1 l

& [d(

i n

N l L

a t

I L Z.

kno 3

IN 'ta A. h

!,l r

b o

3 k..

i te 4a.

1 t A N

=

d' v I_

t l G n Lt S_

lctl.

Li ls.

t t

1 kb s,

b a

h N k k c

t i L

v,

- N.

• L. I I.

8 tiG l'

o N.,.

VN n.

i i ig

,I C,

TP a_

Tb o.

m

- u I

hV s.

C 1t AL t

• aT n I.

u..

- b I

m3 b-h Lt s ul g

1. i i' 3

IiSl i

l'

• ,t h_

L d.

i e6

^

A 0 :.

A 3 I6, t

1 4...

Nv ap s

h l

t.

A a

y W CN

. even though the inspection plan was transmitted to NPC Headquarters several In one instance a PAEC order stopping work in a specific days earlier.

area was not forwarded to the site until PAEC inrepectors called the order At this time there are no ' obvious to the attention of NPC site personnel.

najor ill effects of this cccmunication problem but the potential is disturbing.

Several commendable NPC generally lacks nuclear power experience.

actions have been taken to cope with this problem. For one thing, several midd.le and lower management people were hired from PAEC where they had experience with a research reactor. Also a nu=ber of people from NPC have been and are being given foreign training. The presence of several foreign consul'tants (ESASCO) experienced in quality assurance (QA) is very encouraging.

Despite these efforts, a problem re=ains with regard to the awareness of Both training and consultant support are needed.

NPC top management.

II.S. Maintainine cuafitv Through the Work Interruption.

Evidence Interruption of construction poses several threats to quality.

PAEC so far indicates that reascnable precauticus are being taken.

inspections are specifically aimed at ensudng the centinuation of these precautions.

Major concerns about the interruptica are:

(a) Erected or partially erected systems nay be inadequately protected frem the environnant.

(b) Materials and con:ponents already on-site may be i= properly stored.

(c) Records =ay be lost or damaged.

1729 026

. (d) Quality Control (QC) equipment may not receive proper care.

(e) Key personnel may be lost to other jobs and replacement may be less qualified and/or unfamiliar with the project.

(f) Lowering of morale may lower quality.

The first inspection af ter the work stoppage indicates that the shutdown is proceeding in an orderly manner with due concern to the above For example, the placement of concrete under the contai= ment issues.

vessel (CV) is continuing to ensure that the CV will not be damaged in the event of seismic activity. Also equipment is being properly stored. The QC personnel are being retained to ensure that quality can be maintained when work is resumed.

The potential for ill effect will increase if the work stoppage is extended. Protecting the plant during this period has become a =ajor responsibility of PAEC.

III. PROGRESS AND RECOMfENDATIONS DUR!NG THIS PERIOD In the beginning of thi quarter my efforts were do=inated by a course in the physics of large reactors. This was followed by a review of and reaction to the Three Mile Island (DfI) incident. The public hearings were the principal activity at the end of the quarter.

III.l. Training Course en Physics of Lar2e Reactors This course was part (4 weeks) of the training (12 weeks) provided for tha PNPP-1 cperator trainees. The total program covered a range of It included 9 days topics, frem machenatics to nuclear pcwer plant systems.

en reactor physics fundamentals and 8 days on reactor kinetics.

1729 027

. As no suitable text was available, PAEC asked me to write a text or set of. lecture notes that could be used for future classes as well as the present one. The text was written. It consisted of about 205 pages of text, 127 figures and 37 tables. The contents are listed in Table 1.

This text differs frcm the conventional in that:

(a) It emphasizes PNPP-1 while making detailed comparison to other plants.

(b) It is written specifically for N?P operators (rather than designers, rescarchers, etc.)

(c) It is devoted to large reactors; some knowledge of fundamentals is assumed.

Cnfortunataly, ?AEC was unable to ecpe with the typing and reprcduction of the text. Yest of the material eventually was =ade available to the students but not in a convenient form. Materials were distributed as they were copied, with the Bibiliegraphy ccming first and the other parts fellesing =cre ce less randemly. Chapter 3 was lost and not found until a

=enth or so after the class ended. About half the text was copied in hand-written for=.

Many of the figures and most of the tables were illegible.

Thire is some hope that the text will be compiled in a useful form before I :. eave.

The =atarial was covered in lectures. The course ended with a final examination (Table 2) and a term paper was required frem each studant.

The subject matter was extended to cover va-icus nuclear safety tonics snd te call attention to the problems experienced with the various N??

systems.

!729 028

. Table 1 PHYSICS OF LARCE PRESSURIZFD WATER REACTORS CONTENTS Preface 1.

Introduction 2.

The Reactor Core 3.

Functions and Objectives 4

Reactor Physics Overview 5.

Centrol Rod Reactivity 6.

Soluble Boren Centrol 7.

Mbderator Tenperature Reactivity S.

Fuel Te=perature Reactivity 9.

Total Pcwer Defect 10.

Poisen Effects 11.

Accident Considerations 12.

Analytical Techniq.as Ap pe ndices A.

Pri=ary Coolant ' cop 3.

Cc= version Factors C.

Etndy Facts and Figures D.

Que-tiens 3ibliegraphy I729 029

. Table 2 CORE PHYSICS FINAL EX.01INATION 3 May 1979 1.

Draw a figure depicting the characteristics of a sliding average te=perature program shewing (a) variation of the principal parameters with load, and (b) main ecmponents of the pri=ary and secondary systa=s.

2.

Write the "6 factor" for=ula for the neutron =ultiplication factor and define each term, including" neutren cultiplication factor".

3.

How is each of the 7 quantities in the "6 factor for=ula" affected by (a) an increase in moderator temperature and (b) an increase in burnup?

(Select increase, decrease, ncne or variable.)

4 If the total pcwer defect is 1.5% Jf at 70% power and 2.1%if at 100%

power, hou much must the beren ccncentration be changed to reduce the power fres 100% tc 70%? Assume a boren coefficien: of- -16 pc=/pps and assu=e that other para:ecers, including red position, are unchanged.

5.

If the moderator te=perature ccefficient is -40 pc=/*F, hcw =ust the

=cderator temperature change to accce=odate a power increase frcm 7C". to 100% of full power? Assu=e the power defect of problem 4 and ascu=e that the other para =eters remain ecnstant.

6.

Ecv =uch reactivity would ceed to be inserted to cause the pcuer to increase f cc tero to full power before the Doppler Ccefficient ste::ed the increase? Assu=e that, where ? is percent of full pcwer, the Coppler coefficient is given by:

C() = 0.04 ? -15

.... p cm /".

1729 030

III.2.

Three Mile Island Incident Review PAEC started to receive infor=ation about TMI al=ost i==ediately after it occurred. Interest. ::cunted as the incident became a major

" media event". Af ter a shorr time, the available technical informa-tion became excessive and I was asked to prepare a su==ary for publication in the Journal of the Philippine Radioisotope Society.

The article (Appendix A) was draf ted and submitted to IAEA Headquarters for approval.

The principal cenclusions were that (a) PNPP-1 is' =uch less susceptib.le to such an event than was TMI; (b) enough operator and

=echanical failures could result in da= age to any plant; (c) this incident emphasizes the need for co=petence in the utility operators and =anage=ent personnel.

III.3 Public Hearines My contribution so far has consisted of assisting in the preparation of position papers and in previding support to ?AEC durin;; the hearings.

III.4 Other Activities Assistance was provided in preparing for an in-depth audit of the organi:sticas at the PNPP-1 site. Plans were ec=pleted but the audit was cancelled because of the work stoppage. Inspections centinue to,be =ade to ensure orderly and proper conduct of the shutdcun activities.

Co ants were =ade to the USA Nuclear Regulater; Cc==ission (NRC) about their standards for the qualifications of UPP personnel. The NRC standards are very i=portant because they influence international 1729 031

. standards and the course of actica in developing countries. It was suggested that (a) the present criteria appear dangerously low and (b) the NRC should perform a detailed study and document the bases fcr their standards.

It was receccended that the Philippines try to learn from the TMI incident (a) by having a PAEC representative visit the TMI site and (b) by arranging a loan of NPC persennel to Metropolitan Edison to help with the recovery. The initial response to these reco==endations was negative.

A suggestion was =ade that the US NRC i= prove the value/i= pact statements new being published with proposed new regulatory guides.

This i=provement is needed by the developing countries as well as by the users of these guides in the States.

Infor ation was requested frem IAEA Headquarters cencerning bases for the issues raised by the 1978 IAEA Safety Mission to the Philippines.

Com=ents were provided to PAEC en a proposal by NPC cencerning the PNPP-1 Final Safety Analysis Report (?SAR). The importance of the FSAR was emphasized and it was recc== ended that the FSAR address all issues raised by the most recent versica of the US NRC Regulatory Guide 1.70 (Rev. 3).

A recemmendation was made to NPC that funds be provided so the operator trainess can acquire needed docucents and referenca nacerials while in the US for training.

1729 032

. Correspondence was =aintained with 'ir. W.G. Albert, the Reactor Inspection Expert who is scheduled to arrive in Manila 10 July 1979.

In accordance with my suggestions, Mr. Albert has sent an extensive collection of docu=ents for use during his assign =ent here. These dccunents have been received and catalogued. Several are already being used by PAEC personnel.

Training for senior management personnel of NPC was recoc= ended.

As a = ember of American National Standards Institute Cocnittee N-17, I voted to reject a proposed standard on testing of radiation shields for light water reactors.

Ef forts to obtain documents and reference nacerials was centinued.

The pri=ary source is the US NRC. Doc 2ents received during this period are listed in Table 3.

(This does not include Mr. Albert's docu=en:s. )

Continued support was provided for the Philippine Association for Radiation Protection in preparing for the Asian Regicnal Congress of the laternational Association for Radiation Protection. Preparations are progressing well and a successful =eeting is expected.

1729 033

. Table 3 Docunents Received April - June 1979 NUREG Reports 0017 Calculated Gas & Liquid Releases, ?WR April 1976 C441 DES Gas Hills U Mill Jan. 1979 C472 PWR Rad. Effuent Draft Tech Specs Oct. 1978 0494 DES White Mesa U Project Dac. 1978 0520 SER Construction Palo Verde 4 & 5 Jan. 1979 0523 SER Operation Zi--ar 1 Jan. 1979 0531 Streso Corrosien Cracking - LWR Piping Feb.1979 0544 Eandbeck, Acrony=s Mar. 1979 0550 Rev.1 Standard Distributien Jan. 1979 0560 Feedwater Transients -- B&W LWRs May 1979

UREG/CR Recorts 0125 Fowler etal: Waste 5 Soil Characteristics Jan. 1973 0143 Physics, R. Safety, Quarterly apt.

June 1973 0130 Spinard : Fission Product Heat, QR.

June 1973 0182 Eeddleson: Data, US NPP April 1973 0305 England et al: Decay Heat & ENDF/3 Aug. 1923 02Cl Kla= crus: Preli=., Fire Tests Sept. 1973 0442 Kot, et al: Effects of Air Blast Oct. 1978 Other ANSI /AEME N45.2-1977 Auditing CA Progra=s 1977 ANS 10.5 User Needs in Co=puter Programs

br. 19 79 US Senate deport en Radiatien Health

& Safaty 1978 1729 034

. IV.

PRINCIPAL CORRES?CNDENCE, APRIL - JUNE 1979.

1.

" Proposed Revision 2 to Regulatory Guide 1.8, Persennel Selection and Training," Letter to the US NRC, 2 April 1979.

2.

"Reccccendations Concerning the Three Mile Island Incident," Letter to C.R. Aleta, PAEC, 6 April 1979.

3.

Untitled letter to R.3, Minogue, US NRC, concerning value/ impact statements, 6 April 1979.

4

" Victor E. Andersen's Petitica for Rulemaking, Docket No. PRM-20-13,"

Letter to Sa=uel J.Chilk, US NRC,10 April 1979.

5.

Untitled letter to E. Iansiti, IAEA, requesting infor=ation relevant to the IAEA Safety Mission' issues, 16 April 1979.

6.

"Cc=ents on the 4 May Letter Frem NPC Concerning the FSAR,"

Letter to C.R. Aleta, PAEC, S May 1979.

7.

Untitled letter to Jchn E. Ecdgsen, CNDP, requesting saissic data, 14 April 1979.

S.

" Reference M2terials fer the PNPP-1 Operating Staff," Let:er to Jesue D. Polintan, N?C 15 May 1979.

9.

"Recc=cendatiens for FiF1-1 Rcview 3ased en the Three Mile Island Accident," Letter to C.R. Ale:s, PAEC, 4 June 1979.

10.

"Dacu=ents Received f res Mr. Albert," Letter to C.R. Aleta, ?AEC, 4 June 1979.

11.

"Addi:icnal Docu=ents Frem Mr. Albert," Letter to C.R. Aleta, PAEC, 11 June 1979.

12.

Cn:itled letter to Sa=cel J. Chilk, US NRC, concerning selection and ::sining of NPP persennel, 11 June '379.

13. Untitled letter :o Jchn H. Ecdgson, UNDP, requesting seismic informa-tica, 11 June 1979.

14.

"Reco=endation for Senior Manage =ent Training," Letter to Geronimo Z. Velasco, Philippine Energy Develepten: Board,13 June 1979.

15.

Un:i: led letter to Ralph D. 31anco, NPC requescing inferna:icn abouc the site studies, 15 June 1979.

1729 035

. 16. Untitled letter to D.A. Nethsinghe, IAF.A requesting approval for pu'elication of the til article,14 June 1979.

17.

" Occurrence of a Three Mile Island Accident in PN7?-1," Letter to C.R. Aleta, PAEC, 22 June 1979.

1729 036

. V.

ACCCMPLISEMGTS AND RECCMMD;DATIONS IN PREVICUS PERIODS These were listed in the previous progress report.

VI.

WCRKING RELATIONS I have continued to work closely uith my counterpar:, Dr. C.R. Alata, Chief, Depart =ent of Nuclear Regulations and Safeguards.

Dr. R.J. Falabrica,

. Chief, Nuclear Training Depart =ent and Acting Special Assistant to the Cc==issioner for ?NPP-1 has been away much of :h s period; he was on " rural service" and then en an LAIA assign =ent in Vienna. Contacts with NPC, '11PCO and E2ASCO perscnnel were largely limited to formal meetings prior to the s: art of the hearings.

The major change was the depar:ure of Cc==issicner '.ibrado D. Ibe. The Ccc=issica hac been recuesting early retirmnent or sick leave but neither has been granted. Shortly after issuance of the construction perni: for

?N?2-1, ha reportadly lef t the Philippines and inmigrated to Canada. Later

he sick leave was granted and officially Dr. Ibe continues to be the Cc=nissionar.

Dr. Zoilo M. 3artolone, Deputy Cc== ssion, has been Of ficer-in-Charge cf FAEC since Dr. Ibe's depar:ure. The change has had no disecrnable 2: pact en PAIC's aff ectivenasa as a regulator,' agency.

t.r.-..

.u? u~n.cu r.s.e Not acplicab'.e

~

1729 037

... -.. _,..nu. _r,..

No: applicable

. d.

PROGRAM OF WORK The next quartL wi;L end my..ssign=ent here. Departure in late August is planned. I will then be on. leave until the end of 'the contract.

The remaining time here will be devoted to the following activities:

(a) Assisting ?AEC in the public hearings on the safety of ?NPP-1.

(b) Attempting the complete resolution of the issues raised by the 1978 IAEASafetykission.

(c) Assisting in obtaining the USA export license.

(d) Providing training for NPC =anagement.

(e) Ccmpleting the text on the physics of large reactors.

(f) Cocumenting the PWR systems lectures.

(e) Assisting ?AEC on specific problems as requested.

. GENERAL CONSIOERATIONS AND CONDI"' IONS The regulatory agency continues to function despite underfunding and under staffing. Efficiency is reduced by working conditicus though electrical power failures are less frequent ncu ::ut su=:er is ever.

Transportation is a chronic problem thcugh not as bad as it was last year.

1729 038

Appendix A SUCLEAR INCIDENT AT TFFIE MILE ISU.ND A'D IMPLICATIONS FOR THE PHILIPPINES CHARLES A. WILLIS*

Nuclear Reactor Safety Exper International Acccic Energy Agency / Philippine Atomic Energy Cc= mission AES~L CT The Three Mila Is!and inciden: is revicued and implica:icns for

he Philippine nuclear prograc. re discussed.

It is concluded :ha: the incident can be of subscantial benefi: to :ne Philipe nes if the associated political problems are precerly handled.

Intrcdue:icn The wors a:cident in che 22-yaar nis: cry cf the i.'S civilia. a:ccic ener;y program s:arted a: :,:00 a.m., Jednesday, 23 March 1979. Fra for:ni.3h or more that acciden; was hae line news around :he ::crid. The naus reports, he:cever, conveyed r.cre enati n than inf ernatica. Thus. this repcrt, :,ased en the estarial lu:ed in the bibliegraphy, was prepared :o provide a su=sary of the ple:hcra cf :echnical informa:icn acu available.

=This article is based en :ha au:ncr's

w. e:cpartisa and does ne:

engage the IAEA in any way or imply any c:nnitment en the par: of the LW..

1729 039 090

• o[N y ;

J tL m

m

. In the Philippines, interest is not so zuch in the Three Mile Island-2 (TMI) accident as in the implications of that accident for the Philippine nuclear program. These implications are far frem certain since they depend pri:tarily on human behavior. Thus, any assessment is open to criticism.

.Nevertheless, an evaluation is offered for the reader's consideration.

The Nuclear Power System.

The TMI-2 plant is secewhat like the plant acw being built en the 3acaan Peninsula. The Philippine plant, hcwever, is of, significantly safer design. A nuclear reactor is the source of heat. The heat is transferred frem the reactor to the steam generator (heat exchanger) by water ficwing through a pri=ary ecolant icop. The heat is trcnsferred frem the steam generitar to the turbine by steam. The turbine turns the generator and prcduces electricity (which is the purpose of the plant).

Frem the turbine, the steam ficws into the tendenser where it is changed Inte liquid water. Fro: the condenser the water is returned ta the staas generator by the f eedwater pusps, thus completing the seccndary loop.

At IMI there ara 2 pri=ary and 2 secondary loops. The systen is depicted in Figure 1.

Figure 2 is a ci=plified sche =atic, shcwing the principal cc ponents and ficw paths.

The Accidant.

At appr:xi=stely 4:C0 a.m., the feedwater pu=ps " tripped"

(;cwer to them was cut off) because of the failure cf another component (2 Ocndensate pu=p).

A feedwater pump trip is not a rare event and the systa= vas designed to cope with it.

1729 040

Loss of feedwater requires protactive action because without feedvater the staas generators sccc run dry and lose the ability to renove heat frcs the primary loops. The system is designed to (1) start the cuxiliary feedwater pumps which keep steam generators functioning; (2) control pressure buildup with power operated valves, and (3) procptly shut down (ntrip ) e.ne reactor if necessary.

u The autecatic centrol system started the. cuxiliary feeduatcr pumps just as it shculd. Unfor:una:ely, "the lines f r:c the auniliary feeduater pumps to the stea: generators were blocked by valves which had been cicsed by rain:atance persennel.

Cperating rhe plan: uith the auxiliary feedwater pumps v:17cd cut was the principal caus i the accident and it was an agregicus error, both on the part of :he personnel who cicded '.ne valves and :he opers:crs who centinued :: keep the plant "cn line" in that ccnditi:n. Ih2se acticna were clear violations of the operating license recuirc anta (the technical specifica:icns).

Witheu: feedwater, the steam generator star:Ld :: lose abili:y :o receve haat.

In abou: 1.5 cinucas :he 50s were dry but the inc:runents wrcngly continued := show 10 to 15 inches of water. Thus, much of the energy genera:ed by the reactor had to stay in the reactor and pri=ary Icep. The coolant water temperature and pressure increased rapidly.

In ] :o 5 soc:nds :ho powcr -:porated relief valve (?CRV) cpened because the pressure La the primary systa= uas :co high (2:55 psi). The FORV is intended :o open just as i: did to avcid overprassuri:a:icn of the prinary D "

lD

" l0

' F V sys:c=.

o o M d i

'i 1729 041 m

. Even with the relief valve open, pressure continued to rise and in about 6 = ore seconds the reactor was tripped beccuse of high pressure (2355 psi). The reactor trip system functioned properly and shutdown was accesplished quickly.

In only 3 secends, the primary systec pressure dropped to 2205 psi. At this point, the PORV should have closed but it did not.

Further, thera was no direct way for the operator to know of this failure.

Failure of the PORV to close is the secen1 nachanical fai2are (taking the initia:ing event as the first)'.

There had 'caan at least tuo c:her known instances of this failure (at Oconee and at Davis-Besse) so it should not have been a conplete surprise to the operators.

Failure of the instrumentatien to inform :he op;rators that the PCRV s

stuck open was part desien deficiency in that (1) the design nade no pecvision for a direct indica:icn of an open PCRV and (2) the ther:ccouple tha: should have shewn : hat steam was still flowing threugh che pipe was not properly installed. This failure was aise partly nechanical failure in chat a safety valvc had been " weeping" enough to keep :he ther=ccouple hot even nich the PORV closed. Finally, this f ailura was part cperator err:r in tha: there was enough info reatica available to nake the operatcr aware of the f ailure; afecr 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />, an operatcr f rec the other unit (T:C-1) ca=e in, recognized the clues and closed the blocking valve (which is the backup to the FORV).

Cencinuing blewdown through the open PCRV caused the prinary systen pressure to continue :o drop.

1729 042

.About 4 ninutes after the accident started the pressure fell to 1600 psi and the energency core cooling system frCCS) star:ed to inject _

water into the cere. The ECCS pu=ps had started when the turbine tripped and had reached full speed in 30 seccnds. The high pressure ECCS is intended to handle just such an accident as ;his, where there is a ralatively small break in the pri=ary syste=.

Radioactive decay was heating the primary ecolant water faster than blowdewn through :he open PORV could cool it so the va:er was enpanding.

This, with the addicien of water frcm the ICCS, caused :he pressurizer

0 fil.1 with watar.

Six minutes into the accident the pressuri:er was "sclid" (filled with wa:er, wi:heut a staan bubble).

The injecticn of wa:er into a solid pri=ary systea by the high prescura ECCS can cause ever-pressuri:atien sc the operator =anually

ripped :he ECCS pu=ps.

The first pump was tripped 4.5 ninu:es in:o the acciden: and :he second 6 minutes later. Uithou: :he cold wa:er being injected, tha coclant :: perature star:ed :c risa and 1.5 minutes af ter

he first ECCS pu=p was tripped, the hot leg temperature reached satura:ica and the watar flashed to steam.

Tripping the ECCS pu=ps was :he second najor operater error, caused in par: by bad information.

The operaccrs did not kacw the PCRV was open.

%cra importan:1, they did not knew the auxiliary feedwater system was 7

nct functioning because che tags which showed that the pumps were being na n:sined hid :ha warning 'ights on :he cen:rci pancl.

Ac 4:08 a.m., eight minutes into the accident, the operators opened

he valves tha: isola:ed the auraliary feedwater pumps. This re-established a hea; sink and cculd have saved the day.

In a half-=inute :he coolant 1729 043

te=perature started to drop and in 2 minutes the coolant water shrunk enough to bring the pressuri er level indication back on scale. The operator then re-started the ECCS pu=ps (new in the nor=al ciake up mede) to con::al pressurizer level.

For nearly an hour the situation was quasi-str.bla with the decay haat being re=cved by,:he steam generators and make up injection matching blowdown. Had the operaters realized :he PCRV was open at this time, core damage cculd have been averted but the faces were not sokind$

ne ECCS pu=ps were in use, replacing the water being lost through

ne open PCRV.

"'hese pu=ps were needed to provide cooling water to tha reac:cr coolant (RC) pu=p seals. Also there was concern about vibration of :he RC pu=ps.

To avcid da= aging :he reac:c ccclan: pumpa, the

parators decided to :u. off the pu=ps and relf en natural circulation

a =cve the ccolant be:seen reac:cr and stea generaters.

The systes :..: designed for decay heat r2=cval by natural circulatice.

Hwever, in TMI-tfpe plants na: ural circula:icn works only if tha secondary side of the staa generater is fulu of water. Een the RC pumps were shu:

dcwn :he wa:er level was very Icw, enly 5': of :he cperating level. Even if the steas generator had been full, gas or steam in the lines could have prevented natur:1 circula:icn. There was no natural circulatica so again bicwdevn was :he only heat rat = oval techanis=.

Shuting down the RC pe=ps was the third. ajor operatcr error.

A heatup :ransien: followed.

In 15 =inu:ss the het leg :e:peratures s

vera off-scale, ever 620*F, while the cold lag temperatures gradually dr:pped to the te=pera:ure of the makeup water (150*y). D e ic: leg

a=peratures stayed off scale for scce 3 hcurs.

7in 1729 044 3

D

. e Failure to restart the RC pumps at that point is dif ficult to understand and could be considered the fourth acjor operator error.

At 6:20 a.m. an operator finally realized the FORV was open and closed the (backup) blocking valve. Had this happended before the RC pTmps were tripped, they would have been out of trouble but at this point there was little i==ediate benefit except to force = ore of the

=ake up water into the core.

With blowdown stopped, reactor coolant prassure rose rapidly, frem about 630 psi at 6 :20 a.m. to 2150 psi at 7 :00 a.m.

The blocking valve was then re-opened to prevent over-pressurization and one RC pump was re-started for abou: 15 minutes.

The neut;on data indicate that the cars was pa):ially uncovered f cm 5 :f 0 to 6 :f 0 a.m.

The ini:ial core da= age probably occurred at this time. Certainly fuel was damaged and gases were released.

3y 9 :30 a.m. the operators realized there was a gas bubbla in the top of :he reacter vessel. The bubble was a threat because it could (1) expand and encover the core if pressure were reduced, (2) enter :he pipes and disrupt natural circulatien or (3) da= age the RC pumps if they were operated. Thinking the gas to be s team, the pressure was increased to 2100 psi to collapse the bubble. The attemet failad because the bubble was composed pri=arily of (1) hyd:cgen from retal-water reactions, (2) helium frem the damaged fuel rods and (3) fission prcduct gases, also f cm the da= aged fuel.

At 11:30 a.:. the blocking valve was re-opened to permi: the pressure

o drop so the core ficed tanks could be used and so the residual heat re=cval (RHR) system could be ac:uated. The flood tanks partially dis-1729 045

e

. char;ed but the pressure could not be reduced'below about 420 psi so 'he RER systes could not be used.

The pressure was under 700 psi for 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> and the core probably was uncovered much of this ti=e.

Further da= age during this period is quite likely.

This was the fif th serious operator error.

At 5 :30 p.s. the biceking valve was closed. By S:00 p.m. the pressure returned to 2300 psi and an "a" loop RC pu=p was started.,

A: that poin: the principal crisis was over. The reactor was shut down.

Flow was established. The break in the primary syste= uas closed and :here was a heat sink. Scrious probices re=ained but the likeliheed of a nel:dcun wta virtually elisinated.

Tha core had suffered considerable danage. Most cf the fuel rods ar2 ch ught to be damaged but evidently there was no celting of the fuel.

The cladding =ay have suff ared enough da=aga to ec=plicate its reteval.

A larga amount of radioactivity had been released.

The gas bubble in the er: actor vessel continued to be a cause for concern. Much caref ul :h;ught went in:o devising ways to re=cve the bubble. Ul:izately, the deciaien was made to si= ply let the system run and gradually abscrb tha gases.

The.possibilitf of a hydrogen explosien cen=anded the headlines for a considerable period.

In fac :here seems to have been a local hydrogen axplcsion in the con:ain=an: building a: abcuc 2 :CC 3.m., while :he reac:cr pressure was at a low value.

The explosien raised the centainmen: pressure

23 psig (or =cra) but seems to have dene no damage. There was concern 1729 046

_9_

.5 : hyc:cgen sight build up in the centain=ent building and support an 2p' s :c :ha: w u'd destroy the contain=ent.

D1I has a hydrogen i r.::ne: 5: lack cf proper shielding delayed its use until April 2.

At

c.a : ;;ne the hydrogen contantration was 2.1", well below the fla=mability

.n : f 2*

and the explosion limit of 8".

Thereafter, the hydrogen

-nc A: Lon decreased gradually.

s Iy 29 Ap:11, a n:n:h after the accident, che system actually was

-:.;:s in c.:e ca: ural.irculat;:n :cde.

ne Onta.rmct: sys:an fun::icned effectively :o reduce of f-si:e

..::.. 2ith: ;n ene anali releases that did occar received considerable 4

-.n The ':.:wd:wn through the open PCR7 was to the drain tank.

The

s. cean eas c:t issli ed in the beg;:alng so when he drain :ank go: :c

-"a

n ents were au cmatically pu=pec into -he =iscallanecus s:crage

.t ary b.;1c;ng Unfor:;nately, :hese tanks altsady tere n: 2.:a of the bicwd::n lic,uld ip:11ed in:o cne auxiliary cul cin-

-. - u: rc: 3 2 sis c. Cases that evcived f::= this liquid leaked to :ha 3:s:spnt:e b.:

he releases care quita small, C a r s tr.r.2 t was scla:ad before the sericus fuel damage released lar;e 25..":s c: raalca::ivi:y :: :cntainment, Radia:icn levels inside containmen:

, ;,w : :: 30,000 rad /hr* range but :he peak off-si:a value seems :o have

=r. J 00- rac/nr*

The to al radiation dose :o the pcpulatten w :hin 30 ::les

'.t : :;;;r.g :ne :::s: 3 days of :he ac:ident was calculated :o 1e less :han 12 c:5e these same pecple recalvec in :he same period fr a na: ural Ap ;

13, 721 people who had been wi:hin 3 miles of :he acciden:

.as.

3*

c Je ?O,200 taa/hr, a lathal dose wculd be received in a minuta er ;wc.

E xp,ur - :o C C04 :sd'hr for ene hcur : ght increase an average A=arican's

. hin.s s :: lethal cancer frc= abou: 0.19 :o abou: 0.l?0CCC.,

D " #

lD "D

~ %[ j

^

cu,N cu' I

.sr aL

_a

. were checked with a "whole body counting" system and all were found free of internal contamination from the acciden:.

As a precautionary measure, the pregnant we=en and pre-schcol children were evacuated frem the area within 5 miles of the plant. This evacitation was race== ended two days after the accident started and probably was of no significant value in reducing doses.

Principal Events 1,

The auxiliary feedwa:er pu=ps were valved cut of the system (in violation of the technical specificatiens and ce==en sense) - - operator error.

2.

All the feedwater pu=ps were tripped as a result of a ccndensate penp failure. - - mechanical failure.

3.

The power operated relief valve (?CRV) opened preperly to contr:1 pressure.

The reactor tripped ner: ally.

5.

The FOR" failed : close as it shculd - - =echanical failura.

6.

The instrumentation did act infern the operator that the PCR" was Open. - - compound failure.

7.

The emergency c:re c cling sys:cm (ICCS) properly started injecting va:er into the core.

3.

The operator :urned eff the ICCS. - - cperator error.

9.

The operator res:ar:ed the ECCS pumps, new in the ncr=al nake-up made.

10. The Operator Opened the valves : hat were keeping auxiliary feedwater cu: cf the steas generator.

1729 048

11 -

11.

The operator tripped the reactor ecolant pumps to establish natural circulation even though conditions precluded natural circulation --

operator error.

12 The blocking valve (backup for the PCRV) was finally closed but had to be reopened to prevent everpressurization.

13.

The core was uncevered and unricusly da= aged; gases were released.

14.

The core was re-pressurized by closing the b ; eking valve.

15.

The pressure was lowered in an attempt to use the residual heat re= oval systa= (RHR) - operator error.

16.

The core was further da= aged during this period.

17.

The pressure was coturned to ncrtal and a reacter ecclant pump uns started; this ended the nain crisis af:c: 16 hours1.851852e-4 days <br />0.00444 hours <br />2.645503e-5 weeks <br />6.088e-6 months <br />.

18.

The systa=

tually was placed in the natural circulatica cede nearly a =onth later.

COMPARISON OF TEP22 MILE ISLAND AND PNPP-1 The superficial differences becseen the IMI and PNPP-1 nuclear stacc supply systems (NSSS) are shown in Figures 1 and 3.

Ec.h have two steam generators and one reactor vessel.

The IMI unit has a higher povar rating (880 MWe vs 626 ewe) so the TMI steam generators nust be larger.

IMI has 4 RC punps while PNPP-1 has two.

A critical difference distinguishes tha TMI centrol approach fr:m the

'Jeatinghouse (3) approach in handling foedwater transienta.

Por IMI-cype plants the decision was nade to sacrifice safety =agrin to gain improved system availability. A loss of feedwater does act initiate a reacter trip in a TMI-type plant (as it doe: in a 3,* plant). Rather, the resulting

= Westinghouse 1729 049

. pressure transient in the reactor is centrollad by the PORV so that if the secondary system recovers, the TMI reactor need not be shut down.

This i= proves availability and economics. Ecwever, delaying the reactor trip deposits a great deal of extra energy in the core.

The steam generator (SG) water capacity of a TMI-type plant is relatively small, perhaps 20~ of the capacity of a hi SG.

Thus, if f eed-water is lost, the SG gces dry very quickly in a TMI plant.

The TMI pressurizer also is relatively snall so it has less capacity to handle surges and transients.

The design for natural circulatica is relatively poor in TMI-type plants largely because of the relative elevatiens of the reactor and the SC (see yigures.1 and 3). Also the TMI piping provides a dangercus apper-tunity for gas blockage of natural circulation. The bend in tha IMI hot leg which brings the flow dcwn into the SG is both the highest and the hottast point in a pri=ary Icep (except the pressuriter).

It foll:ws that gas is most itkal'y to form'in that bend.

In :entrast, in a 1 system the high point is half-way through the steam generator, where the water is such cooler and a gas bubble is less likely to f'orm.

In bcth systems, the intent is that any gas bubble is less likely to for=.

In both sys ta=s, the intent is that any gas bubble should be formed in the pressuriter where it dces no harm. This werks less well in the Ei! system because the surge line between the hot leg and the pressuri:er is' relarively 1:ng and crooked (Figure 1).

The P'TPP-1 ECCS pu=ps are not the regular charging pu: ps so use of the ICCS dces not incarfare with the use cf the RC pumps.

1729 050

Thus, PNPP-1 is such less susceptible than TMI to cora damage frem a feedwater loss incident. Of course, mechanical failures and cperator errors could overceme any degree of censervatism in design.

The Regulaterv Role Like the Philippines, the US has regulations requiring i==ediate notification cf the regulatory agency in the avent cf a serious accident.

Initially, the operators did not feel that the events justified

'--adiate no:iff.catien. By 7:45 a.m.,

it was clear that an energency situatien existed and the Nuclear Regulatory Co==i,sion's Regional Office was called. The NRC's answering service had seme trouble centact-ing the Of ficer-in-C..arge bu: centac: vas made and an emer3ency teac tras c. son en its way.

The regional office is in King-of-Prussia, Pa.,

only about an hour's drive free the site. By 10:05 a.m. the NRC team was at III.

The initial team probably was ccmposed af health physicists and other e=ergency specialists. Ecwever, by 30 March the NRC had 30 technical anc 3 public relatiens people en si:e along vi h sc=e upper =anage=,en:

personnel.

The centributiens of the NRC to coping with the proble are not clear.

It is obvious that the NRC could not have reached the site in time :c help in the early stages.

.t.lso the NRC presence did not prevent the nalepera:icns which left the core uncevered frc= about noon to about 3:20 p.m.

The public ce==unicatiens proble= vas cc= plicated by apparently 1729 051

t.

ccnflicting stacaments frca che NRC and frem the utility. The evacuation of young children and expectant mothers was based on an NRC recccmendation.

The NRC people could have been invaluable had there been serious contaminatica proble=s out these did not develop.

The NRC =ay be criticited for not analycing more fully cecidents of this sort. However, had the plant been cperated in accordance with the license require =ents the accident would never have happened.

A f ull-ti=e resident inspector sight have prevented the accident by ccsuring ccepliance with the technical specificaticas.

The NRC =ust be given credit for acting quickly to prevent si=ilar occur cuces in other plants and for disse inating tachnical infor:stien ab out the event. The NRC also played a ajor role (along with other government agencies) in the environmental conicering effort.

?tincical lessons 1.

The safety systa=s can protect the public even af ter gross failurcs and =aloperaciens.

2.

The operating staff cust cope with accidents and abnornal ccnditions; generally cutside help will cece too late and vill have too little k culedge of the specific system to be of grea: value.

Ccusequently, the operations ocaff needs i= proved training, aspecially

= ore ti=e en the si=ulator.

3.

Utility nanage:cnt nust have an effectivo system cc ensure tha:

the rules (tech. specs. etc.) are obeyed. Generally, utility manage =ent 3

needs training in their safety respcnsibilities.

1729 052

. 4 To effectively cope with an accident :hera is a need for:

a) Effective, realistic emergency prepara:icns including access control, quality assurance, on-and-of f-site cce=unications, emergency supplies and agreements with local authorities; b) A good computer system for transient analyses; c) An automatic control room legging system; and d)

I= proved recording and priat-cut systens (especially a high speed printer for the alar =s recorder).

5.

General purpose instrumentation is needed :o previde information in abnormal si:uations; this includes closed circuit TV and microphones in the reactor building and other key areas.

6.

In areas which nay con: sin racicactive =c:crials, such as the auxiliary building, ven ilation should be fully cen: rolled.

7.

Ins:ru=entation was deficient in several areas at 1MI and shculd be re-assessed at all plan:s; exs=ples include lack of flew =easurements frem the auxiliary feedwater pumps, lack of information about the ?CRV sta:us and radiation =ccitoring instru=en stion with insuf ficient range.

8.

Radioac:ive waste handling capacity may be eco s=all.

I=tlica: ices f or :he Philiccines The Philippine nuc1 Jar project will be saf er because of the TMI accident.

The principal benefits are 1.

Utili:y management will be scre aware of their respensibilities.

2.

The opera crs, who are just beginning their training, have been nade acutaly aware of their respcesibilities.

1729 053

. 3.

The PNPP-1 systems will be reviewed and := proved to correct any tefi:lencies that are recognized as a result of TM1.

e.

Because of the bed publicity, nuclear sales are likely to be 11:::ed for a while; this =ay provide opportunities for purchasing uranium cad possibly the second nuclear plant at bargain prices.

The unfavorable implications of IMI for the Philippines are equally

eal Perhaps the nest important are:

The poli:: cal cli= ate in the USA is likely to be such that serious 2

d.tia;u;:les may be experienced in obtaining export licenses.

2.

The active an:1-nutlear organi:ations will use TM1 to prc=ote publ:: fear and oppcsi:icn to nuclear developmen:s.

3.

S::e pecple r.; be:cce sincerely :encarned about nuclear raae:::

s :s:y. Jespite :he perfect pub 1 : safety re:Ord.

Shouls anti-nuclear at:::udes become ::o streng, key persennel s.

may be disc:uraged and even sw :ch :o other fields.

5_2:ar r

!he recen: accident a: Three Mile 1.aland vas the worst in the his:ory

f :he USA nuclear pcwer prcgras.

5: 11, no ene was killed or even injured.

The radia:icn doses from the accident to :he pecple in :he area were less that the deses fr = nature to these same pecple in the same period of time.

The acciden: has i=pertan:

plica:icns for the ?tilippine nuclear p:: gram.

.aere are ac h good and bad aspe::s :: these i=plicati:ns but n:h eff e::ive Philippine leadership, the acvantages can cu:veigh :he

_ acrantages.

o**

• o 9 1729 054

. JB t W==

_m.

BIBLIOGRAPHY 1.

US State Dept. Cables, C. Vance to Distribution, almost daily,1 April to 29 April 1979 and less frequent thereafter.

2.

"USNRC Preli=inary Notice of Event or Unusual Occurrence", I'NO 79-67, 28 March 1979 through PNO 79-67 AR, 4 June 1979.

3.

" Incident at nree Mile Island", USNRC IE Bulletin No. 79-05,1 April 1979.

4

" Incident at Three Mile Island, Supplement", USN2C II Bulletin No.79-05A, 5 April 1979.

5.

" Incident at *hree Mile Island, Supple =ent", USNRC I2 Eulletin No. 79-05 B, 21 April 1979.

6.

" Review of Operational Errors and Systen Misalign=enta Identified During the Three Mile Island !acident", LSNRC IE Zulletin 79-C6,11 April 1979.

7.

" Review of Opera:icnal...", USURC Sulletins 79-C6 A & 3,14 April 1979.

3. "Raview of dpers icnal...", USNRC Sulletin 79-06 A Rev. 1, 1S April 1979.

9.

" Events Relevant to 3o4. ling Water Reactors Iden:ified During the Sree Mile Island Incident", USNRC II 3ulletin 79-08,14 April 1979, 10.

UFNRC News Release,

','01. 5, ';o. 5, Week Ending 5 April 1979.

11.

Hendree, Jcseph M., Chair an, NRC, Let:e :o Idrund G. 3rewn, Jr., Governce, C21ifornia, 5 April 1979.

12.

"The Ordeal at Three Mile Island", Nuclear "evs, Specal Report, 6 April 1979.

13.

"Nucicar Night:are", Ti.e, 9 April 1979.

14.

" Nuclear Accident", Newsweek. 9 April 1979.

15. Marshall, Eliot, "The Crisis ac hree Mile Island: Nuclear Risks Are Recensidered", Science, 204, 4389, pp. 152-3 (13 April 1979).

16. Carter, luther J., " Political Falleu: Frec *hree Mile Island", Science, 204, 4389, pp.154-5, (13 April 1979).

17.

Carben, Max, Chair =an, Advisory Cce=ittee en 2eactor Safeggards, letter :o Victor Gilinsky, NRC Cem=issioner, 20 April 1979, 18.

"Open Meeting, 2riefing en :he Principal Fac: ors Related to Current S:atus of Operating Plan:s", 23 April 1979, (unrevieue' -- - - 'pt of NRC ree:ing)

Ace-Federal Repc :ers, !=c.

1729 055

11 C. Michaelacn; Rencval During a Very Snall Ereak LCCA for a 19.

3 6 W 205 Fuel", Unpublished Papar, Tennessee Vallay Authority, Jar Sotts, M1. V. Lebe, Energy Incocperatad, Briefing en the Three 20.

Mile Island Irfay 1979.

Marshall, Elicg the Danage at *hrae Mile Island", Science, 21.

Vol. 284, No. )4-496 (11 May 1979).

22.. "22Sch Geners:(Unreviewed transcrip: of :he 5 April 1979 neecing of the Car.nittee en Resc:or Saf:.;ur:s), Aca-Federn1 Raporters, Incn, D.C.

of reedtester Transients in Pres-23.

"St2ff Repor: :ic Assessnen:

surized Water signed by :he 3abecci and L'ilcc.-: Cenpany",

USTRC Repor: May 1979.

24 "crshall, Elicainary Repor: en Tc.rse :ile ulcnd, ji:.2nce, Vol. 204, No.30-231 (20 April 19 7M.

25.

Car:ar, Jir.ny,'s Connission en the Acciden: at Thraa 'sile Isl:nd," (E::2c) 'lhite 3cusa, kashin:; :n, O.C.,1.7 Apr il 197 9.

26.

"Public Mee:ia: ion cf Oriefing en ?rincipia 7sc:crs Ralacia; tc Current Staa:ing 21r...:c", Enra ria-scd tr nceript ci :::C Inc.

naa: ins, 23 Apcc Fedar:1 :'.aporters, 5:atuc cf Operatinz 2 esc:xs",

27.

"Ccatinua:io. 2 elated :c Curran:

(Cnrevictied :: NRC nee:ing 27 Apr:1 '.977? f.ca rederal ?. aper:ers, Inc.

J.G. Eerbein,.a Edisen Co., le::er to US SEC, A:en. ii. Dent:n, 28.

ransnit ing ? sequence of area.:s at 2C,16 April IP.

29. Laurence Scarn risis: "hree Mile Island," ';cshing::n ?cs:

l'.eport, L'ashi.:1979.

O 1

P,DlOr

,o' 'jf

'u n

3 1729 056

e5 N

Pa C

<a 7 =

&s 9

p

' ~

i

.c $

a; q wMw

$$i d il h'

M. E

+w

a

o..

a s.*

.x w m.

o a

4

\\

3

[

l-

-~

- y o

g y x

.a L

4

}

w' 3

9

<)

9 l

i 2

',\\yshT i WgsO / )

5

) T, 7

I

. I t::f *3:.

' W,'

3 ' \\'t'&l j

• J 7'j?]c'-

s-j ~... n :

'd m,

/, '.,\\

}s.L i

v w

/

v g

n /

)6 ff

!)

f i!

l[

ll

'i l

i :.

8 I le l'l

!'.M, %*x~% l:l.l C _ y 1

2, 5

53( (s l

u s

~

a w

a N

N h

k

,/

's' E.

=

o w

'^

j%+:

a

,/

n

\\j ii,/

,g

~

1,r~

W

~ ?. ' ""\\~\\ ta,u \\, W-y;)

<\\ ;\\

\\\\

1

\\a.

/.

aM 1

D [\\

N h (D.d l i C..I. [, i T,,, h\\k / :\\

IS N.

\\

W I

). d@

4 1i I.

i' id f /.,'.;i *,

i,i, \\

'e.

i

~.U/ ';;Q j;E ;.! p..'lL J!

',\\

j,l j ':1, b

IK

~l

 :'s s

j d.a -

1

=,

l

=

a 22'

- _ 's.m,

'a

.e

,~

.\\

J

\\,[NT "

s 6i C

I \\\\

'i C M

. %.s

. -..- 5.b, m

\\q!-

r

• s L

a

. rl l3

)

~

[ _!

j. J.,..

9 F' i

4 l '

. r: a<

g.-

{,

p

\\op E, 'lp} ~ ~ ~ ~' ~

a u

.Q. '

L o

r *, L ' i, O

w 8

s. -

..o' o\\

e

\\ Va

)

g

.i

-n

...s

(,.i v

i J

C.. r' l N _,' '. i 9

• )

=

n

~ ' g,

,i 1j-S 38 8

f c !!

. ; f,'

J R '

C.

'l~.

I

-- _ 3.* !,r:

P p :. =..

i\\

i

[

cs k:'

\\

t N

~y j J'.f,

.e.

s

/

=

I

p. /

p 4

and G3 i

e w

4 b

I ses f%=

a

~

g a

a a

e M

4 m

o 2

e_

a, e

'ee i

G. C 4

s em M

=

tem 1

Tovg Tn,Tc 2

V PRESSURlZER REACTCR CONTROL 4

o 9 STEM Thrctile

]

PRESCWE SENS2;G k

in Th OEVICE hPs M'-

~r s y TURBl"E REACTOR Ts ca Primcry g'

gga Secandert CONCENSER l Loop Locp NEUTRON C

CETECICR U C m8 i

Recctor Ccch.nt

-*"C'"

Fmmp nmp p

3 i

CCnCEi! SATE f

g STC'i;tCE TNM h,Rei m '2G N L@qect:cn Nap (EC... O]

F Auxii.Jy IcedWCf3T

\\

g, REFUELE WATER I

STORAGE TleM f

l scam 1c cua22 ce ta s.uum St=e 'was FIG. 2 D F F' D "D'

~hf'"

\\ b Ji 2 5 ~ l $

1729 058

i 1

TCR

{

l PRESSURtZER A

REACTC6 CCCL%if V

n ia).'JP -,

\\

\\

v lb v h' p

Gg

~ H q

l

'}

}

J T.

,/

/]

\\L REAcitR PNPP I T'30-LOOP REACTOR COOLANT SYSi251 I

b b

1729 059

Appendix B Reconnendation for F"PP-1 Review Based on the Three Mile Island Accident As a result.of the accident at Three Mile Island (UI) unit 2, the safety of reactors generally is being reassessed. It is reco: mended that

'PAEC re-review PNPP-1 safety in light of what has been learned frca Sil.

The DiI accident was the combined result ef design deficiencies, s

mechanical failures and operator errors. There are lessions to be learned and inprovenents to be =ade in each of these areas.

Desian Deficiencies At VfI a number of deficiencies beccme evident, particularly in the instrumentation area. The accident was worsened by the fact that the operators did not know that:

1. t!ie auxiliary feedvater punps were valved cut and were net delivering water to the staan generators.

2. the power operated relief val. e was stuck open,

3. reactor coolant (RC) inventory was lov (despite a high level in the pressurizer), and

4. XC saturation temperature was exceeded.

Many of the design deficiencias were unique to TMI cr at 1 cast to BG plants. These need not concern us.

Bovever, certain of the shortemhss exist or may exist in PHPP-1 and should receive attention. S.e principal issues is this category seem to be:

1. '4 hat should be done to improve instrunentation?

"oP]a ID~ o A

1mm 1729 060 e

3 2. Can the residual heat removal (RHR) system be used safely with the RC highly contaminated?

3. How can the control system be upgraded?

4. How can control room design be improved to enhance operator performance?

5. Can an on-line diagnostic computer be prcvided to i= prove operator performance?

6. Can general purpose information system such as audio systems and closed circuit TV be provided to aid operators in abnor=al situations?

"cuiement Malfunctions Equip =ent failures both triggered and ecspounded :he tiI accident. These failures suggest a need for improved reliability.

If practicable, acticn should be taicen to:

1. decrease the frequency of feedwater transients,

2. i= prove the reliability of the power operated relief valve (2CRV),

and

3. improve instrument (pressuricar level, etc.) reliability under accident cocditions.

Oserator/ Procedural Shortcemine Without operator error, there would have been no DfI accident. Operating the plant with the auxiliary feedwater pu=ps valved out was a gross error.

Failure to realize the PCRV had not closed was an understandable error

ensidering the instrument failure. The near-fatal errors of tripping and D *

• l0 D

T M h

_a o M o

AA=

1729 06I

3 failing to re-start the RC pumps when the hot leg temperature went off scale is al=ost beyond comprehension. The list could be extended.

This experience indicates a need for improved operator capability.

In my judgment the first group of NPC personnel to undergo operator training have the basic abilities needed.

It remains for us to ensure that they:

1. receive sign.ficantly better, training than was given the TMI crew and,

2. are provided werking cceditions that will saximite performance The =ost evident 19 needed ir.prevements in operator training are for inproved understanding of the system and more time en the simulater.

The question of proper working conditiens is more ccmplen. Clearly it includes the following:

1. improved e=ergency procedures;

2. better accident analyses, particularly (a) analysas of very small break LCCAs (b) realistic transient analyses and (c) a detailed failure mode and effects analysis for the plant;

3. backup capability for quiely analyzing transients resulting frem varicus =alfunctions and/or operator acticus;

4. a continuing education and evaluatica program to ensure that operator capability is maintained;

5. recognition and rewards ce=mensu$ata with the respensibilities; and

6. recognition by YPC top manage =ent of the vital role the utility plays in the reliability and safety of a nuclear plant.

D rE r 70 P D W i

> MN m &

1729 062

~

_ Imolementation The foregoing discussion provides a general outline of what needs to be done. Specific implementation will require further consideration.

It is reco= ended that a series of =eetings be held with NPC/ti!PCO to discuss the various issues and to provide a basis for for=al requirements.

Prcept attention to these matters is important since the sooner changes are made the less costly and more effective they can be.

1729 063