2CAN058704, Requests Approval to Utilize Temper Bead Weld Repair Technique,Described in Winter 1985 Addenda of 1983 Edition of Section III of ASME Boiler & Pressure Vessel Code for Temporary Repair of Pressurizer.Supporting Info Encl
| ML20214E883 | |
| Person / Time | |
|---|---|
| Site: | Arkansas Nuclear  |
| Issue date: | 05/12/1987 |
| From: | Tison Campbell ARKANSAS POWER & LIGHT CO. |
| To: | Murley T NRC OFFICE OF ADMINISTRATION & RESOURCES MANAGEMENT (ARM), Office of Nuclear Reactor Regulation |
| References | |
| 2CAN058704, 2CAN58704, NUDOCS 8705220297 | |
| Download: ML20214E883 (43) | |
Text
ARKANGAS POWER & UGHT COMPANY CAPIT0L TOWER BUILDING /P. O. BOX 551/tlifti ROCK, ARKANSAS 72203/(501) 377 3525 f, ctNC CAMPuttL v4 n u a nt blear Operat onis 2CAN058704 U. S. Nuclear Regulatory Commission Document Control Desk Washington, DC 20555 A
ATIN:
Mr. Thomas E. Murley, Ofrec.'or Offico of Nuclear Reactor Hogulat;un
SUBJECT:
Arkansas Nuclear 000 - Unit 2 Docket No. 50-36.
Licenso No. NPf-6 ANO-2 Pressurizer lleator Repair 10CfR50.55a(a)(3)_ Request
Dear Hr. Murley:
In accordance with 10CfR50.55a(n)(3) AP&L requests the NRC's approval to utfilzo the Temper Ocad Wold Repair Technique as described in paragraph NB 4622.9 of the Winter 1985 Addonda of the 1983 Edition of Section Ill of the A5ME Uoller and Prossuro Vossel Code for the temporary repair of the ANO-2 pressurizer.
On April 22, 1987 a containment entry was modo on ANO-2 to assess the potential of a small Icak on one of the Safoty Injection tanks as it.dicated by a higher than normal make-up rato to tho tank.
Whilo in llo containment, a general walkdown of the Roactor Coolant System (RCS) was conducted.
During tho walkdown, leakayo in the area of the pressurizar was detect.l.
Additional entrios were mado on April 23 and 24 to further ovaluato the leakayo.
from thoso entries, it was dolormined that two small leaks existod, ono from the high point vont lino located over the pressurizar
(-100 dpm) and another originating from the pressurizor lower head (~60 dpm).
ANO 2 was shutdown on April 24, 1907 por Technical Specifications and a 10CIR50./2 report was mado assuming that the lower head leak was in the pressuro boundary.
following insulation removal and visual Inspection of the pressurizer lower head, the leakago source was dolormined to be the X1 pressurizer heator sloovo, iho pressurizer lower head was drained by removing the C1 heater.
Attempts woro then mado to remove the heator elements from the X1 slenvo at, well as two other heators which had failed 070D200?97 070D12
}0 PDH ADUCK 0D000:160 P
PDH g \\
v.v.nv.oa. ournu men n,.,eo
l May 11, 1987 (AA1 and T4).
The AA1 heater was easily removed, however, the XI and T4 heaters could not be extracted. A video camera was inserted into the l
adjacent AA1 heater sleeve in order to determine why the X1 heater could not be removed.
The video from this inspection revealed that the X1 heater had ruptured and damaged the X1 heater sleeve.
The upper head pressurizer manway was then opened for visual inspection and damage was also observed to the T4 heater.
An adjacent heater (Y4) was removed to assess T4's damage.
This inspection revealed similar damage to that of the X1 heater.
- However, the T4 rupture had not caused sufficient sleeve damage to cause leakage.
j Repair of the XI and T4 heater penetrations will require the removal of the I
existing sleeves and the insertion of temporary plugs.
Additionally, weld l
repair of the corroded area adjacent to the X1 heater sleeve, resulting from the RCS leakage around the X1 heater, is desirable.
This area is approximately lls Inch in circumference and 3/4 inch deep.
In order to permanently repair the vessel in accordance with currently l
approved ASME codos, as specified in 10CFR50.55a, with current tooling limitations, the repair process would require the removal of all 96 heaters and personnel entering into the high radiation areas inside the pressurizer (10 R/hr Gamma, 12 R/hr Beta) to remove the heater support plates as shown in Figure 1.
During the heater removal process, which requires the heaters to be cut out, the heaters may be damaged to an extent where they could not be used again.
Preliminary estimates for new replacement heaters is 22 weeks.
If a sufficient number of heaters are not available following the repair the unit could not be returned to operation.
Following the support l
plate removal, a corrosion resistent repair could only be accomplished by sending personnel into the pressurizer to perform the weld repair.
l Additiona11y, the exterior of the vessel may also require welding to provide sufficient strength.
If this were necessary, an elevated temperature post I
weld heat treatment would be required.
This heat treatment could only be l
accomplished by placing heater blankets inside the pressurizer on the lower head wall.
The above described repair would result in excessive radiation exposure to the workers performing the repair, a prolonged outage for the repair (potentially up to 22 weeks), and considerable expense to AP&L and its customers.
The Winter 1985 Addenda of Section !!! of the ASME Code allows Temper Dead Welds to a depth of 1/3 the vessel wall thickness without l
performing a post weld elevated temperature heat treatment, which is required by earlier versions of the Section !!! code for weld depth of greater than 3/8 inch.
This additional depth is necessary for the vessel repair.
The Temper Dead Weld Repair Technique, described in the Winter 1985 Addenda of the A5ME Code, provides an equally safe and quality alternative to the unusually difficult method described above.
The fact that this technique is endorsed by the ASME supports AP&L's position that it would bo l
equally safe and of high quality.
Additionally, it is our understanding that the NRC staff has reviewed this later Addenda and found it to be acceptable; however, it has not yet boon of ficially endorsed in 10CFR50.55a.
1 The proposed repair technique was described in detall in a meeting with ths NRC staff on May 8, 1987 in Washington D. C. and at the ANO site.
It j
consists of cutting and boring out the damaged sleeves so that there are no remaining flaws in the vessel (verified via NDE) and in the case of the X1 I
l l
__________________ _________ _ ___ _ _ May 11, 1987 penetration, grinding out the corrosion damaged area on the exterior of the vessel.
The cavity will then be mapped and examined by NDE.
A close tolerance plug, made of compatible SA-533, Gr, B, CL.1 (P3) material, will then be preheated along with the weld repair area on the vessel in accordance with the ASME Code.
Following preheat the plug and the area it is to be welded to on the pressurizer will be " buttered" using E8018G electrodes.
The plug will then be inserted into the sleeve hole as shown in Figure 2.
The plug and cavity will be welded using E8018G electrodes and a qualified welding procedure.
Progressive MT will be utilized to assure weld quality.
The welded area will then be heat treated for 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> (450 to 550
'F which will not require heater blankets be placed in the vessel).
Finally, the weld area will be ground to its original contour and examined by MT and UT after a 48 hour5.555556e-4 days <br />0.0133 hours <br />7.936508e-5 weeks <br />1.8264e-5 months <br /> hold period at ambient temperature.
Prior to making the repairs, the entire repair process will be performed on a full scale mock-up with obstructions present, including wearing anti-Cs per the Radiation Work Permit.
This mock-up training is required per the Code as part of the welder qualification and is one of several more stringent requirements of the more recent Code.
Other requirements include:
- The weld procedure qualification test assembly to be of the same P-number and group number, including PWHT at equivalent time and temperature as the vessel material.
- Dimensional requirements on the weld procedure qualification test assembly (i.e., depth of cavity, thickness, area around cavity, etc.)
- Minimum 4-hour PWHT holding time for P-3 material after weld repair.
l The design of the plug is in accordance with paragraph NB-3222.4(d) of the ASME Codes.
A stress calculation in support of the design is presented in Attachment A.
Because the plug and the walls of the bored out sleeve penetration are made of non-clad SA-533 Grade 8 material, corrosion is of concern.
Independent l
assessments of the affects of corrosion for the remainder of the current l
fuel cycle (approximately 8 months) were conducted.
These evaluations considered aerated and deaerated conditions and included:
i
- General corrosion of exposed P3 steel
- Galvanic corrosion of P3 steel coupled to clad Inconel
- Crevice corrosion of P3 plug and pressurizer base metal l
- Stress corrosion cracking and hydrogen embrittlement of P3 material l
The results of both studies were consistent, showing corrosion rates less i
than 2 MPi under deaerated conditions and " MPY under aerated conditions.
The evaluations concluded these amounts of corrosion to be acceptable for the remainder of this cycle.
Qualification past this cycle of operation will be further justified, if necessary, to allow for development of tooling to implement permanent repairs.
The two corrosion studies are provided in Attachment D.
L
_ _ _ _ _ _ _ _ _ _ _ - _ - _ _ _ _ _ _ _ _ _ May 11, 1987 Finally, AP&L will make every effort to remove the damaged X1 and T4 heater parts from the pressurizer.
However, because of obstructions which may block their recovery, it may not be possible to remove every part. Because of the relative tranquil environment in the pressurizer and the fact that the surge line protrudes several inches into the pressurizer and is covered by a screen, it is highly unlikely that any parts left in the pressurizer could migrate into the main coolant loops. An analysis in support of this contention and which evaluates other effects of the heater parts being left in the pressurizer is provided in Attachment C.
As committed in the April 8, 1987 meeting, AP&L will provide an estimate as to the size and quantity of parts left in the pressurizer when our recovery efforts are complete.
In summary, it is AP&L's position that the proposed use of the more recent code case for the repair of the ANO-2 pressurizer would provide an equally acceptable level of quality and safety to those repairs made using currently-approved codes specified in 10CFR50.55a.
Utilizing the more recent codes will reduce down time, cost, and radiation exposure to the workers involved with the repair.
Very truly yours, g/4sse
<t.
T. Gene Campbell TGC/0EJ/sg Attachments i
h m.
T LE
/
I I
I S
/
(,
I S
E V
R g
E f
Z I
1 R
E M
U RU
\\
S G
I F
S E
l g
I_
R P
g RE h
W N'
I
. xb O
L E
E T
T A
A L
L R
R P P
E E
P h
U U
S S
2.0 00
+.010' DIA.
.000"
.7 5' MIN.
(_ _ _ _ _ 1 h
n
- 4.5" 4
+ 0.2813'
+
MIN. WALL
[l 1
1.375" 1.938'
.0156" DI A.
REF. O.D.
FIGURE 2 I:MSTALLATIO:M OF PLUG g
TYPICAL DIMENSIONS
ATTACHMENT A STRESS CALCULATION PRS-200 N
~
(
bSNTYMS NUCLEAR POWER SYSTEMS COMBUSTION ENGINEERING, INC.
s '
COMPONENTS ENGINEERING Chattanooga, Tennessee s-see.
HEATER PENETRATION REPAIR O
PRS-200 N/A CALCULATION NO.
CONTRACT NO.
ARKANSAS NUCLEAR ONE UNIT 2 PRESSURIZER CUSTOMER COMPONENT D
w PRINCIPAL ENGINEER o /d~/87
~
~
~
/
PREPARED TITLE DATs M
DESIGN ENGINEER f /.5' d 7 VERIFIED TITLE DATE A STRUCTURAL SUPERVISOR
[ f [7
'R'EVIEWED TITLE DATE
~
i
_d
\\
/ APPROVED TITLE DATE L. -
A-W_/be
'~-/- 8'?)
L
- ar# o oate COMBUST 10NJENGINEERING o-F satt
/
M Yf/77 COMPONENT ENGINEERING curcaro care cominnet no A//A HTLE NG~ATCC 550v'61~2A *rro& ECPA r2 Au oo3sete-Aees e c 6e' zacou r W
Pee ce=G =dA&-
crerdc cAc.c.we. A v =ad rn't i:H:7c
~
or e=:
drocCEwed7~ fbR AOJAC.Ed7~
HE~A 73rK
~ 't v~45 W i m A 7 r o W ovECLAP.
Cow:cce Y.40 40-OACGd7~ NEA 7ER PEWE7X?A7~ odG l
- Ce'dC -tDc24
\\
N f
A
\\
1
~
\\- f. sc /
I.is t 4----
A; f, = 0.e,c Ae, e
A,,
A,,
Q
~~~~
7 7,l/ /'/
/
a
' ~ ~ ~
//",
/ //E v
l-,
L p q-(,
C. 3 *7 S~ "
i
[Acca c n r.a--
f.
So 7~MA 7*
Ap,
- 8 2, Aa, =
^ ; ' (p.s pa)
O
- t. 37 9
=W
=
As, - (4,
'* ')(i. ve?)
f L, = 6, -> (c,
'fL")(i. <e9.) = /. e n.wa
\\
d.,
==
/. S~o i
'7'isE*eCF o e G~
La = 5~376 - (_,
=
- 3. e 74
tJHrcH rc Le n rwa d 7Hs c.=H. r or ROUf0CCCM'O]7"* G'7"A 7'CO rd 7WE" CoOC.
l
&a = Ag
/. 879. d *
=
(
i
% = (i.,- 4'.J6.<n)=- (am)O.m) 4.?! 8 rd *
~=~
R EY.
DATE BY CHECK R EV.
DATE BY CHECK REV.
DATE BY CHECK
CCMBUSTICO ENGIZERING war
?
o, e
te 7,',,;,7,,,,, NJ,,"
$ypfg r/5(f7 couponenr encineesina TITI.E
(
//CATCG 5t/Er2A r:0 ^!
06"PA%2.
Auouinece AerA ca-Ga=woour (Coar 'o)
To YHE-PCCM'r'3CA 8t c Ag[A of Gerdocuy As r3 G.tved SY g= A,, - Ag
= 4. 9 3 8 - /. 3 ? 9
,/
e
= 8. S b'~ f Z~W 4/WCeC
~rHE~
Ci/TreC A ee"A As
-:=
al.=, e://
YNG" ARE"A A r,y
-TNoc Jd A80 Ve~l GA r==,
t.):TH:d 2874" or THE~ CD/rcec=dd" of rae-PedCr2A7:Od MO WO Ho2G" 7NAd
/. Y 9 9 "
O C E P.
MCA cu erHedrc reon rNC F:C e-o
.z~d ozc A T C'
(
THA T A
Gerdocu ~r or
/. O OCPrH ado
- o. iec ~
cC.verv sevAcear ro rue uoenc 60er
=c e E q u.r e c o.
5'o A = 6. oX.7. isc) g
- a. ise es ' <
- e. cs,.ra "
=
7NE~ 4dr.rCZPoArEC G2 VOO u r
' races roe r,A &.G.
.z w ACLCpr*
(\\.
REV.
DATE BY CHECK REV.
DATE BY CHECK REV.
DATE BY CHECK
C"-/ - 2 7, I
- ==
6 oa's
' COMBUSTION ENGINEERING 's-er' 5
c.
hh,7 COMPONENT ENGINEERING c[,,,c,,,
f4 e
c,,,
TITLE h'CA TCC S /e~T C A 7"noe/ SEPACR N/A >< rHu r f &dev;eA 7:es/ l): A H e.sez Q Ps e rHC o2:G Zr/A 4 CALcu LA 7*:0 T N T L. : H : 'r $
W EE ^/f~ORCE H E^)7~
22 A OJA cc^)7~ MATER A 1*^/ c r e A 7 o M c O Vc el.A p.
Coa crocre 7u-JO ADJ A CG^/7~
HGA Te;~4 P 6tJ C 722A 7": cs/C
( d N G ~2 6 ~
d,ypa x, 7WC Or/G~
.rc 60eEO Ot4 7" 70 t-/A x =H u q O =A ME"7ER cdH:C.W MCETC 7W[
AG2A 2C:s/s=~OCC5"WOf7~
E Q u 2 2 G"H 5~^) M v
- CE*+/C-lD2 4 l
%L
- i. x, i -
W l
A
\\
\\N d*
2.57c: '
-A e, A-x 2
e a
v h.
f e
/
N t.499 t
(
L A
C315"
.Z~r-k]AC P.eEY:Ou Gl Y T H Q 4)^l 7~NA 7~ kg
- bd, 4/HC+/
- l., =~ i. S*C /
- 2. s v 4
- 7ueecroce-
&., = c e,s -i. s c i
=
ata=cH rs U=rs=^t rue coor c =H=r.
7nesj W
Ag, =-
y (c. Syc)
A,, = (2.e74 d* k)(i.199)
'.199 (7. 748-o%e )
Ae, = Aq =e:>
- 7.3vc dm
=
n
- 3. e 7 s %,, =
//. c i <
2.997 "
Q d,y
=
m To e s/E~ PCr/GYRA r o rf CA A/ Se~ e/LA eG e~o ro D. 99 7 "
ora m ies~e.
R EV.
DATE BY CHECK REV.
DATE BY CHECK REV.
DATE BY CHECK
no W
s~-4-87 i
n s-<ir 4
o, 6
COMBUSTION ENGINEERING t aa aco can c' caro COMPONENT ENGINEERING
[,[,((c,,,,.
~ "
,fj4 TITLE NCAr62
&dereA r=orf ECPA R E~VA L u A T Od O C~
kl Loy'- 6~ a. Cl.
Co
- t. 37s~
Mrd C _--
a, o
0.0 MhW
//
/
i b4o
.2d~ Coa /cc I
(
h Pee. sc
/~
3A - S"33, Ge 8, C4. /
% = D C.~? K s =
I 3.0
/
P-3 talet.o
/
-t
. Pc.ac
/
'S'A -C33, Ge e, Cc. /
02 E~tpus VA L.cgr S.8?c' 3',., c W. ~7 k 3r
.2~r-rc GHot.d&
r WA r" Th'E" Plug Ge~cHerev' AeoVC HEui4
~7-Ne~
2G'q uzeEHEM o f*
de - 30 0D. 4 Cd) or ryG" coca^.
7De-WeAo pedevaA rzod cad eG" G're so 70 ace r nic-26Qu.=ecmedrc cc-W8-8820.
7Ds:egr ces,
f%eAGeAPH d6-833/ Ca) e d:sueer THA7~ 7?/s" W e - 3 0 a s. /, N e ~ s p ai..o,
(
eEqu.=ecuedm-oc-W8-302/. 3 Ado d6-300;?. -?
4e 6~ H e~7~
FOR-PeEWi'u2T LDAOZA/G.
REV.
DATE BY CHECK REV.
DATE BY CHECK R EV.
DATE BY CHECK
r,v Sj'4/6Y COMBUSTION ENGINEERING ! sarev 5"
o-6 v io one gh(d [S[' (,'2 COMPONENT ENGINEERING
[,,,',", c m j,j TITLE
?
4' a rce PedereAr=bd 2CPA XR c
E~vs t u A -r o d o #-
Auov'- Trccc Pc u G P uc c J A c c.
rwzcxWes5,7~
t
=~5 Oe~rce r-r:dCO A/8-3/33,3 koc P = D.z~ a:'cr c5e o,
/.3,s" D
/_. ::h' O. c'- "
l
- ' S 1.80 (u t c~ D. o )
O,
/.375 r r7-7~ w s376" G=
/0.
~7~
c"Sor-1 r:c. V//-//00-/ op-Te c rzorf I T k=.OQ c~eort fr~ s. v// - //o /-S oc' Tec-r:oe/f 6 =c.20 ht =
7"i/c+/
- 6 W#O E p =.
- p. cc es:-
=
3("/r)
=co)
(
a, > e - o.s ec.c o<
W 7~. =, /3 75
.Oce?S "(d.A.)=,Do "
man
);~o e C7
= 0. O "
~
o L
- S
/. 0 5" (use /. 4)
=
=
c?,
- o. o
/_ r7-7=
, Coo" Oo
=-
/o.
7-f~eo q
):~xg V// - //co- /
of JCc.r~ orff A==.OS 5204 5r~ s V's/ - // o / - S or J e c r.= a d.~l2 C 6 = Do. S ecr 77/EV 4 tS W3 8)==
0, ~7/ k'G r
):"7
=r w
3('Vr) 3(/o.)
P, >
a =
p.6~
ecr or
(
N[
~7"
=
- Q o o +. O d DS(c. 4.)==, Od'D 5" mo REV.
DATE BY CHECK REV.
DATE BY CHECK REV.
DATE l BY CHECK i
S/YU COMBU5 TION ENGINEERING sarer 6
o, 6
a -eso a.re I/f/(,3
,',,",c,,,
[
~
COMPONENT ENGINEERING j
TITLE
(
McA tee W e A rz vov! ADE~PArg EvAcuArrod oc-ALLOY- &cce Pcuc (Codr'O) 7u~c NEAO-Pcus 4/CLO Ur&4.
gg-Jurrzs~:'Co
.1 v=~oc fAr:Gue~
S Y' JNo s.J rdG' 7//A r THE RE-
+
?l
- Qu
- :cEMGd7Y Cf W8-30DO. % (o')C)-(4 ) Aes~ Mci.
W e-8D00. < CoX/)
A rwos pMs erc.
n 5'cevrcc~
Petscuee Cyce. E".
bec.=nco Cyc&a 5- (Ws )
//CA nn Yb Coo e.cowd 5oo Cvcccc
//Yoeo 37~AWC.
rt:.q 1 -
/O CYCLS T f rA c
'n i
.700 CYccc3 Tora c A f = ~7/o CYcct3 g
AccouJABLC 0YCc.EC [h! )
A Tm =
QG. ?
cc.:
Q = 33m Bo./
J:Y.c
(
\\
F"e~ ow fr~ s Z - 9. /
7~4'a~
Ac cocJAeLe-MuHeE~e ok~
6YCc E8
=G'
/V =
l/00 C.YC L6~3 4
Als = 7/o d,q
= //0 0 7 H~ e e's qu. ernedr.cc He 7-W6-Spoo. 4 (c/)Co)
WoeaAc Scav:ce-pemuee 5~w'c. r u A r=oJz,
k wxr=cAdr Peescuee fi~uc.-ruA r:o4 (6P) sP=
Lp 3
3 Gm 4Heec=~~
Q= D6~O0 Mr Y
= pc. ?
esr m
a
/a. s est (r~ e 2~ 9. / @ /o cyccsti-)
S-r
=
2 41 } =
d590 PG.r REV.l DATE l BY CHECK l REV. l DATE l BY l CHECK l REV. l DATE l BY l CHECK
W&nts A/~ 4lar@7 L
Mio COMBUSTION) ENGINEERING ; swer e
7 on B MM N,[
COMPONENT ENGINEERING
- )[,,],,,,
[g TITEE McA nre CevereA v=M Ecm=e Au.ov - Ercen Rug (coar'o)
EVAC.us7:oJ or 9 *'
Me-Sopo.4cdXo) Coar'o Co aorr xoJ Pa=4 Occuecaces
- 1 CR=)
Okr.)
RAdr bodofdG Padr UdcoA O JG
/0 % TrcP.Z~dcecAcE-235o 2/So to #
/0 Po Crep CGra cAce'-
WoeHA C.
VAA. A *T"IoW RCACYDe 7e~rp locc or Scot.A d 7 f[c M D/So
/ ~700
<80 dox or doAo d6'A e YW DDS~o 2250 Goo
//yOeo TA T.CC Q~r 2/ R$~
3/ Q$
/O A
E':Gd:F":CAdr~ M ueG" #7.uC 7uA r.;"od [R.- )
.zw dP = A P c
em A/ user
@ = 250o PC.r J
Cd. ~7 E T. -
=
m g
5" =
/2. S~ k'.7 (&G Z~-9. / @ /o C Yct Eis".)
~
6 P ~ 3 (2 5oo) a'#*b J
Bo P s'.r
=
c
- c. 7 f~ec~r rns' Aeo v8~ 7~A et.G~
77/ces~ Aes-48o CYCCES of T.~G^/.r r.r @ d r P2Mu ed5" f7 M C.Tu A YZQ ^/C :
7kE'
^1A X
^-/ur.f p2 meg ~
~
fz.ac -ruA -rroJ
.:rc dP S'Q 5~ - / ~7c o
/425~ Mr
=
=
mu
'7~ c AccouJABCG~
P2MueG~ f2*aCTuA Y ~orf a
r; dP=
4 P i
4 3
o s,
n i
duceE-j p = Q S~o o p c.:
O REV.
DATE BY CHECK R EV.
DATE 8Y CHECK REV.
DATE l BY CHECK i
Wh' nu 5~l1/2;7 L
=#eo oara COMBUSTIONJENGINEERING wen F
or 6
he r/rg,7 COMPONENT ENGINEERING c[,,,c,.
fg TITLE YCA ~1*G~2 YG~~72A *1'%o!)
SCAAZQ EVAf uAT Ot/ Or bu.DY-ETCCL E uG ([DWY'O)
L M e-Sopo.-1(d M o) coar 'o.
G ',,== D C.~7 K::~r Q = /09 est (kz~ e. E-9. / @ 180 cvccG3-)
G~o 4 P = 3 (2Soo) 2C. 7
= 3-stoc ps r A
dP MDS d P, 34co p 3*=
=
=
THC eCQu:QEHCMr 13' HE"'7~
de - 200o. 2(d)(3) 7e~ acceA TVGG' SZfX~C&GWCC
- S ~r n e r v o
-7 0 Durowd TMG~
plug ado MCA O efMA ="d r3o 7WC2 MAL.
Oue:JG A4c.
~r2 A d J.=~E d 7C',
vWC2Cfb 2 E'
(
TN:J
eCQurcEMedr 1r Me <.
W@~ 3220. 4 (c')(4) 7CHPCRA ~ rue 2~ Ch*fWRGhfCE-
~
AfoenA t-Ecev=cc~
l
%c pcus AMO NcAO ee/7Ard rCo7'HEeHAL.
ouezw A u-.
reads cdrs, 7wme fc2e;-
7'N:t i
EGpurec ME~d7~ =;s' MC7~.
//8-2000. s co'XS) 7En~1PEeA 7v&G~ O A escWc.E -
Ozst=HrtAe h fA T E G = A G 5' 7uc nus, scA o Ado
+/ce-o M 4 v s ~ 7 W C G'A H E '
do cs;~r~~rc= cd7~
o f' 7WE~e'WA L GWPAA/croJ, races fDee' rn.zz-esyurecues/r ::s' HG7' il We-3p o o.<(dXc )
PfecaA uxcA c.
foA cc-(
rzec e Aee-do wecyAdrcA c.
t oAO5' QW acc.o, mene,oec 7 w z e e e 9 u = a c.-
PfCd7*
.;C3' M C tK REV.
DATE BY CHECK R E V.
DATE BY CHECK REV.
DATE SY CHECK
ATTACHMENT B CORROSION STUDIES A-MCC-87-002
int:rCffl30Corre ponden30 9
&ONIBGETION ENGINEERSISS
(
TO: '.J. A. Walker FROM:
J. F. Hall ec:
K. R. Craig May 1, 1987
)
W
,.. J. Heilker - Chatt.
- M. D. Turnmire - Chatt.
A-MCC-87-002 R. R. Mills J. F. Fox Corrosion of SA-533 Grade B Class 1 Steel at ANO-2 Attached is a analysis of the extent of corrosion expected to pressurizer and plug material exposed to borated water at Arkansas-2.
, (.
i I
l J. F. Hall JFH/b Attach.
l I
O 9
CORROSION OF SA-533 GRADE 8 CLASS 1 STEEL AT ANO-2
1.0 INTRODUCTION
-SA-533 Grade 8 Class 1 steel will be exposed to borated water as a result of temporary repairs to the pressurizer at Arkansas Nuclear One Unit 2 (ANO-2).
When used as pressure boundary materials in the primary system, low alloy steels such as SA-533 Grade B are clad with stainless steel to isolate the materials from the primary coolant, thereby minimizing corrosion and corrosion product generation.
However, temporary leak repairs at ANO-2 will result in non-clad j
material in a pressurizer well and in a plug being exposed to primary coolant
[_
for the remainder of the current fuel cycle (approximately 8 months).
The
'k objective of this review is to determine and document the expected corrosion for SA-533 Grade B steel and low alloy steel weld metal.
2.0 LITERATURE DATA Reference (1) presents the only literature data that are applicable to ANO-2.
During the 1965 refueling outage at Yankee Rowe, inspections discovered two j
small areas where the reactor vessel cladding had been breached.
The cladding defects were mechanically produced after a surveillance capsule became loose and dumped mechanical test specimens and other debris into the lower head of the reactor vessel.
The lower head base metal was A302B steel which had been clad with Type 304 stainless steel plate using an intermittent spot welding technique.
This fabrication process resulted in large areas of the cladding-base metal interface being unbounded.
Thus a breach in the cladding l
would expose large areas of the underlying base metal to the reactor coolant.
l
(. !
I
Because of concern about corrosion and hydrogen embrittlement,, the reactor 4
vendor initiated a test program to determine corrosion rates for A3028 steel under operating and shutdown conditions and to determine if hydrogen absorption was a concern.
In this program, specimens of A302B were exposed to aerated and deaerated solutions of boric acid (2000-2500 ppm B) at temperatures ranging from 70'F to 500'F.
The specimens included both electrically insulated coupons of A3028 and coupons electrically grounded to Type 304 stainless steel to assess Iga 1vanic. effects. Most of the testing was at low temperatures (70 to140'F)in 8
2500 ppe B aerated and deaerated solutions for up to 121 days.
For these conditions, the test program well characterized the corrosion rates of A3028.
The r,eactor vendor did conduct a few short term (6-14 days) tests at 300 - 500 F in deaerated soluticns containing 2000 ppe B.
The results of the study showed that hydrogen absorption was not a concern and that the exposed A3028 steel would corrode uniformly at a rate of about 3 mpy (mils per year) which was sufficiently low to be of no concern over the remaining 25 year lifetime of the plant.
C-E re-examined all of the Reference (1) data in developing estimated corrosion
(
rates applicable to the ANO-2 pressurizer application.
3.0 CORROSION RATE FOR PRESSURIZER MATERIAL, PLUG AND WELOMENT 3.1 Assumotions C-E used four basic assumptions in developing a corrosion rate applicable to ANO-2. They were as follows:
1.
The corrosion of SA-533 Grade B Class I steel in boric acid solutions is equivalent to that of A3028.
Both grades are low alloy steels but 1
there are minor compositional differences, with the most significant being higher N1 in the SA-533 Grade B Class 1 material. The presence of Ni will not adversely affect corrosion.
)
!k 1
2.
The weld material has a comparable chemical compositio1 to the
~
pressurizer and plug materials.
i.
3.
During the remainder of the cycle, ANO-2 will operate 80 percent of the time.
4..
While operating, the coolant will be deaerated.
While shut down, the coolant will be aerated.
.3.2 Rates for Hiah Temnerature doeration
. ANO-2 primary coolant temperature during operation is approximately 600'F.
The maximum test temperature included in the Reference (1) program was 500'F.
Data were obtained for six specimens of A302B after only one week of testing in a refreshed autoclave with a 2000 ppm B solution. For these specimens, the average corrosion rate was 0.6 mpy.
For the specimen with the greatest observed weight loss, the corrosion rate was 1.0 mpy.
These values for corrosion rates, which are based on short term tests, are
(
higher than would have been obtained had test times been significantly longer.
Carbon and low alloy steels usually follow a logarithmic or parabolic corrosion rate law.
The time required to reach a steady-state condition may exceed 100 days.
Rates during the initial transient stage (which includes seven days) will be high compared to the steady-state corrosion rates.
t There are no corrosion rate data for low alloy steels in borated water at temperatures above 500'F.
However, the Reference (1) program did include tests at 300'F and 400"F in deaerated borated water.
These tests showed decreasing corrosion rates with increasing temperature.
Based on this, C-E judge that the (steady state) corrosion rate at 600'F will not be greater than the rate at 500'F.
In addition, the available data are from tests with B levels of 2000 ppm, significantly greater than expected at ANO-2 during the current fuel cycle, and, thus, they are conservative with respect to actual conditions.
l( :
l 1
J 3.3 Rates for low Temoerature Conditions 4
I During any unscheduled outage, the material may be exposed to low temperature aerated borated water.
Reference (1) presented data for tests 0
at 70, 100 and 140 F in aerated water with 2500 ppm B.
For this analysis, 0
100 F was assumed to be of coolant temperatures during cold shutdown. The tests continued for 121 days with several interim weight loss determinations which showed that steady state rates had been obtained.
For the Reference (1) tests, the average corrosion rate for 100'F was 7.0 mpy with an upper bound of the data indicating a worst case corrosion rate of 7.9 mpy.
The non-clad pressurizer material will be coupled to the stainless steel clad in the pressurizer.
The use of dissimilar metals will cause some concern about galvanic effects.
In the Reference (1) tests, the A3028 specimens were both insulated from and connected to Type 304 stainless steel, which chemically is similar to the stainless steel used for cladding of the primary system.
No difference in corrosion rates for the two types
(
of specimens were observed and thus galvanic corrosion should not be a problem.
The limited corrosion that will occur will not be a localized form of corrosion (deeppitting,stresscorrosioncracking,etc.). C-E experienced in aerated borated water indicates that the corrosion morphology will consist of the superposition of shallow pits with no deep penetrations.
Reference (1) also indicated that corrosion would be uniform with no localized forms of corrosion being observed.
l 3.5 Overall Corrosion Rates 1
Based on a split of 80 percent hot operations and 20 percent cold shutdown, corrosion rates for the pressurizer material at ANO-2 were determined as follows using the rates described above:
" Worst case" corrosion rate - 1.0 mpy x 8/12 yr. x 0.8 +
7.9 mpy x 8/12 yr. x 0.2 - 1.6 mils.
l
t
- d 70101 1
- d 91:11 40/20/50 46C6-682CCO2) 33 W0as I
j 4.0 REC 0petEN0ATION For analysis purposes, the recommended value of corrosion for the pressurizer material, plug and weld metal is 5 mils.
This value provides a conservative estimate of the corrosion expected for the remainder of the current fuel cycle.
~5.0 REFERENCE 1
,, 1
'
- bsorption of Corrosion Hydrogen by A3028 Steel at 70'F to 500'F,"
WCAP-7099, December 1, 1967.
(
i t
t '
4 9
e i
(-
m,
.J,
.. d
~'d s.
aw v.uo.s <
NM M"
. veemca mwy ENGINEERING INFORMATION RECORD Document identifier 51 11d55000 T tie __ Evaluation of Interim Recair of ANO-2 Pressurizer PREPARED BY:
IEWED BY:
Name 5 H.CAem.t.
k E.KAc m _
Name Signatur,JRC*wttl bate 5b Hn Signature b 8 M SS Date #/
f Technical Manager Statement: Initials /M Revieweris Independent.
Remarks:
1.0 Purposa This document is to provide the metallurgical justification for an int repair of corrosion damage erta to the ANO-2 pressurtzer head adjacent to a pressurtzer heater.
2.0 Bacireround on April 24, 1987, a nontsolable leak of ~0.02 spe was detected at one of the 96 heater elements in the ANO-2 pressurizer.
The plant was subsequently shut down to investigate the cause of the leak and determine the extent o damage.
The heater elements penetrate the lower head which is fabricated from Grade Be Class 1, MnMcNi steel plate with Inconel weld clad on the inner surface.
The pressure boundary attachment is formed by (1) an Inconel sleeve and a J-groove weld at the junction of the sleeve, cladding and pressurtzer shell and (2) a weld between the sleeve and the heater element below the pressurtzer head (refer to Figure 1).
is between the sleeve and The most likely leak patn the I.D. of the hole in the icwer head.
could This occur if a defect was present sleeve to shell seld.
in either the sleeve or the J groove Scric acid corrosion was found on the 0.D.' of the head surrounding the heater sleeve.
The corroston occurred in the form of an enlargement of the annulus between the sleeve and pressurizer shell and cavity 1 1/2 inch diameter by 3/4 inch deep about 1/2 inch from the sleeve.
Figure 1 shows a sketch of this corroston damage.
I 4
p,, I at 16
51-1168554-00 Prae 2 CF 16 Boric actd cormston has been observed in a number of operating plants.
The most recent occurrences are the ANO-1 HPI line grooving and Turkey Point 4 reactor vessel head service structure damage.I'2 This type of corrosion damage occurs when carbon or low alloy steel components are exposed to leaking reactor coolant.
Alternate wetting and drying produces low pH conditions which in combination with an air atmosphere can result in a very high corrosion rate (- 1 In/ year).
The corrosion rate is greatest at temperatures between 200 and 350F; although, a signif f cant rate is achievable at higher t empe ratu res.
Evaporative cooling of exposed components, as the coolant flashes to steam, can increase the corrwston rate of items that are normally at temperatures where the rate would be low.
The observed corrosion damage on the ANO-2 pressurizer is not surprising considering previous operating plant experience.
3.0 Prenamed Interim Renafe It is anticipated that repair of the corrosion damage and reinstallatton of a new pressurizer heater would require significant effort to qualify and pe rf o rm.
An interim repair is, therefore, proposed which involves the follow f ng:
1.
Wol d repair of corroston damage, t f necessary, with low alloy steel weld metal (P3 to P3) to restore minimum wall thickness and design features j
2.
Remachining the bore l
3.
Inserttng a low alloy steel (P3) plug with an interference fit 4.
Welding the plug to the 0.0. of the pressurtzer snell
{
This approach results in an extremely tight crevice consisting of Inconel clad and icw alloy steel plug at its entrance and low alloy steel between the clad and the structural weld (- 3 1/2 in).
This crwice may be exposed to reactor coolant at both ambient and operating temperature.
- However, ingress of reactor coolant into the crevice is expected to be minimal.
i' k
- =
51-1168554-00 P,ME 3 CF 16 4.0 Mitarfalt and 'Cereenton Consideratf eu Based on a materials assessment of the proposed repair, several corroston concerns may be postulated. They are (1) general corrosion of the exposed low alloy steel, (2) galvanic corrosion of low alloy s. eel coupled to Inconel clad, (3 ) crevice corrusion of the icw alloy steel plug and pressurtzer vessel base wtal, (4) stress corrosion cracking of the low alloy steel, and (5) hydrogen embrtttlement of the low alloy steel.
The pressurizer eny tronment and corrosion concerns are discussed in the following sections.
4.1 Pranturbar Environnent The pressurizer contains primary water which is covered by a nitrogen blanket during most plant conditions.
Typical reactor vessel water ch emis try pa rameters for primary water during shutdown, subcritical and ce t rical reactor operation are shewn in Table 1.3,4 Of pa rticular importance are oxygen and boron levels and tempera ture.
These parameters would be identical for the pressurizer except during shutdown conditions when the pressurtzer remains under a nitrogen overpressure.
This nitrogen overpressure will limit oxygen content to that of operatrng conditions (f.o.
< 0.1 ppm).
Haever, some plant maintenance, e.g. R. C. pump seal repair, requires draia!ng the RCS.
Under these condittons the reactor coolant is aerated.
This condition is not expected to occur untti tne next scheduled outage in 1984.
However, for the purposes of this evaluation, it is assumed that such a condition could exist.
~c..:
i g;;,.
- -E.3-iE: ::E!
- sH -ci 51-1168554-00 PN1E 4 0F L6 4.2 Ganaral Corronf on of I ew A11ev staal General corrosion may be definec as uniform deterioratton of a metal surface by chemical or electrochemical reaction with the envircnment.
Nickel alloys such as the Inconel cladding are generally immune to general corrosion in a PWR environment, t.ew alloy stools, however may be attacked depending on the actual environment.
The general corroston rates of low alloy steels during aerated and deaerated cor.ccions will be discussed in this section.
Many investigators have reported corroston rates of carbon and low alloy steels in various environments.S-IS One particular evaluation was performed after reactor vessel cladding camage occurred at the Yankee Rowe reactor in 1965 due to mechanical camage by loose pa rts.16,17 A reactor vessel surveillance capsule broke away releasing Charpy impact test specimens and other doorts in the lower heac.
Several small areas of cladding were worn away exposing icw alloy steel.
This cladding had been attachec to the steel plates by spot welding.
The total area cf base plate exposed to reactor coolant was approximately 2 square inches; however,10 square feet may have been exposed in a crev fce under the spot weld clad.
A t5crough evaluation of the corroston conce rns tnat might arise as a result of continued operation with exposed base plate was perfcrmed to avoic replacing the clad.
These concerns, the same as outlined in Section a.0, Materials anc Cerrosion Considerations, were resc1ved.
The results of the Yankee Rowe evaluation are presented in the following sections, together w f th the results of other corroborative laboratory results.
Corresion Ratas tlndar Dnamented conditions General corrosion tests of ASTM A 302 Grade 8 MnMo steel (a icw alloy steel) in borated PWR water under aerated and deaerated conditions were performed in support of the Yankee Rowe inctdent.16'17 Th1s data is applicable to the ANO-2 material because several investigators have shown that in the same
m,:. :
-s.:- u -ig: ::55
=ii -:-
l l
I 51-1166554-00 PAGE 5 0F 16 b
environment, carbon and low alloy steels would have similar corrosion rates.6 7,8,12 At deaerated conditions during reactor operatton, the corrosion rate depends on temperature, velocity of fluid, boric acid concentration, and time.5 The influence of flow velocity on the corroston rate of low alloy steel in a PWR (deaerated) environment is shown below.16 Flow Velocity Corrosion Rate f t/ =e MPY 36 3.0 20 1.4-1.5 2
0.2 In the pressurizer, a very low flow velocity is expected; a flow velocity of 2 ft/see would result in corroston rates (from data presented previously) of 0.2 mpy. This corrosion rate for deaerated conditf(ns is significantly less than reported corrosion rates for aerated conditions (See next section for aerated corrosion data).
Other investigators have determined the ef fect of oxygen on the corroston rate of icw alloy and carbon steels.8 10,12 They found that, at ordinary temperatures in neutral or near-neutral water, dissolved oxygen is necessary for appreciatie corrosion of icw alloy and carbon steels.
In water cooled nuclear reactors, water is disassociated raciolytically to produce oxygen (among other species).
Mcwever, the hydrogen (overpressure) in the reactor vessel ( nit rogen in the pressurizer) recombines with the oxygen so that little or no free oxygen remains in the water.10 The corrosion rate of Icw alloy and carbon steels in a deaerated FWR environment is, therefore, expected to be negligible.
Experience with the Yankee Rowe reactor which has operated successfully for 20 years (under deaerated conditions and aerated conditions during normal shutdowns) supports this observation.
l
51-1168554-00 Pra 6 0F 16 Corrusion Istaa Linder Aarated Coeditfang As a result of the Yankee Rowe incident, general corrosi steel were performed in aerated PWR water to simulate s on tests of A 3028 The investigators showed that corroston rates in simulated PW ons.16 ppe boren - stattar to shutcown conditions) water (2500 after about three weeks of exposure at 70F to 140F, attained a s rate of 1500 mg/dm2-mo during shutdown conditions was pr A general correston corroston rate would result This stis) during a two month shutdown.in a rate of approximately 0.0015 inch (1.5 The effect of temperature on the corroston rate is signtfi vertfted through laboratory tests on carbon cant and has been steels.
Sabcock and Wticox reported general corroston rates of carbon steels in aerated 15 solution (1730 ppm borun) at 200F.15 c acid showed corroston rates of Submerged SA212 carbon steel plates 2450-2620 mg/ cm.mo.
2 At 75F other investigators showed general corrosion rates of 1050 mg/dm -mo, less th 2
the higher temperature.1 a eat Other investigators have reported that the corrusion rate in increases up to - 176F and an open vessel then decreases at the boiling point.
indicates that the higher corroston rates obtained by B&W Th is at 200F; however, the lower rates at 70-140F would be more pe would be expected the MO-2 spectal shutdown concittons.
uring S u==am of L aw A11ev Staal General enrrenten Ouring deterated conditions (typical pressurtzer environment) rates < 1 mpy of the low alloy steel are expected.
During aerated corroston conditions (special shutdowns), corroston rates of 9 apy may be p l
l l
__..__._,_.--.-...m.._---
_.r.__,
_-,__,-,._r,__
_--.__r
3
- n
.. ;-s a :g:
- n
- g -;.
51-1168554-00 PAGE 7 CF 16 4.3 th1 van t e cor roit on of L o= A11ov steel caueled to incenal clad Galvante corrosion may occur when two different metals in contact are exposed to a conductive solution.
The larger the potential dif fe rence between two metals, the greater the likelihood of galvanic cormston.
Low alloy steel is more anodic than Inconel and would, therefore, be subject to galvanic attack when coupled and exposed to reactor coolant.
In support of the Yankee Rowe damage where low alloy steel was galvanically cou pled to austenttic stainless steel due to cladding damage, corroston tests were conducted with A 302 Grade B coupons and A 302 Grade B coupons coupled to Type 304SS exposed to simulated PWR envirunnent under deaerated conditions.16,17 Both coupons exhibited similar corrusion rates. Low alloy steel coupons coupled to Inconal 600 would probably setbit less corrosion since Inconel is less cathodic to low alloy steels than stainless steel (based on relattve positions in the electromotive series).
Other investigators have reported corroston test results for salvante corroston between carbon and low all oy steels and stainless steels.5 Exposure of alloy steels coupled (weldec) to stainless steels was carried out in near neutral high purity water for 1000 hours0.0116 days <br />0.278 hours <br />0.00165 weeks <br />3.805e-4 months <br /> at about 546F in steam, steam-water, saturated water, and subcooled water containing from 0.1 ppm oxygen to 15 ppm oxygen and 2 ppm hydrogen in the steam and steam-water mixture.
The chloride content of the water was less than 0.1 ppm. Visual, macroscopic, and metallographic examinations revealed no traces of galvanic, selective, or accelerated corrosion in or near weld zones of any of the welded specimens.
Other investigators also found no evidence of galvanic corrosion when they tested 5% chromium steel and type 304 stainless steel couples at 500F for 85 days in water containing 35 ppm oxygen or 98 days in water containing 530 ppe oxygen.14 y$
e=e
51-1168554-00 PAGE 8 0F 16 In s t.mma ry, galvanic corrosion should not be a concern at ANO-2 during typical operation.
General co rro s ton rates (presented previously for aerated conditicns) were obtained using coupled samples; therefore, th is corroston rate takes into account any galvanic effects present under aerated conditions.
4.4 Crwice carreden Crevice cormston is probably the most important consideration for this repair cosign.
The environmental condittens in a crevice can, with time, become aggressive and cause local cormston.
The Babcock and Wticox Company (B&W) has conducted several stucies on the environmental ef fects of a crevice condition in a PWR environment, particularly in the OTSG tube-tube sheet crev f ee.10'19 Studies conducted in 1967 and 1981 are directly appitcable to the u nde rstanding of the crev ice environment in the pressurizer repair situation, because the materials are similar, and the environment at the tube-tube sheet crevice is aerated during shutdown.
One study conducted in 1967 examined the ef fect of crevices on deposit production at the tube-tube sheet inte rf ace.10 A mockup was modified by drflitng holes in the low alloy tube sheet perpendicular to the holes for seating tne Inconel 600 tubes.
Several holes were used to simulate intermittent borated water leakage in the crevice, and a total of 1050 hours0.0122 days <br />0.292 hours <br />0.00174 weeks <br />3.99525e-4 months <br /> of testing was accumulatea.
The primary deposit observed at the " borated water leakage" locations was f ron oxide as Fe3 4 and/or Fe2 3 The only 0
O other constituents present in concentrations greater than 1% were silica, nfekel, baron, manganese and ch romium.
In most cases, the deposit was confined to a few inches of depth at the top of the crevice (possibly self-plugging).
The rolled-in area of the Inconel tubes contained very little deposit, and the tube-tube sheet crevices examined revealed no cracks after deposit removal.
-,,, - - -, - ~, - -. - - - -, -, - -
51-1168554-00 PVsE 9 OF 16 In 1981, S&W conducted an examination on an intentionally "crey tcod" 19 model bofler.
In this instance the Inconel tube to Inconel clad low alloy tube sheet weld was in tentionally damaged (cracked) and a crev tco was introduced permitting primary water leakage.19 This model boiler operated for approximately 6500 hours0.0752 days <br />1.806 hours <br />0.0107 weeks <br />0.00247 months <br />.
Test results indicated that deposit butidup was indeed occurring in the crevice; the crevice was actually healing itself 1.e., plugging with deposit buildup.
Leakage was detected during the first 1000 hours0.0116 days <br />0.278 hours <br />0.00165 weeks <br />3.805e-4 months <br /> of operatton.
After this time, however, leak testing by helium water bubble and Leak-Tec methods indicated that the crevice (leak path) had become plugged with deposit buildup.
S11ght changes in leakage rates were detected after the system was shutdown for an extended period of time.
At operating conditions these leaks appeared to seal.
No evidence of corrosion was observed on the Inconel weld clad, and deposits similar to those observed during the 1967 test were present on the low alloy steel base metal.
These test results indicate that iron extde formation, the result of low alley tube sheet corrosion in the tube-tube sheet crevice, plugged the leak path of the intentionally damaged tube.
Under ambient conditions the oxtde was less tightly packed and the leak path remained slightly open.
Uncer operating conditions of high temperature and pressure, the oxide became more tightly packed and appeared to more tfgntly plug the crevice.
Other investigators have performed operiments to determine the crevice corrosion rates of low alloy steel.10,16 These results indicate that crevice corrosion rates under aerated conditions would be less than the general cornaston rate.
Corrosion rates under deaerated conditions would be e t nt ual due to the lack of oxygen in th e e nvi ro nmen t.
Additional operational expertences of the PWR and Naval reactor programs show that crevice correston is not normally a problem in PWR systems.13
51-1168554-00 PAGE 10 CF 16 -
4.5 strans carmsten crackina i
Stress corrosion cracking (SCC) can only occur when three conditions are presents (1) a tensile stress, (2) an environment conducive to SCC, and (3) a susceptible mkterial.
Inconel clad would not normally be susceptible to stress corrosion cracking in FWR primary coolant.
Stress cormston cracking of carbon and low alloy steels is not expected to be a problem under BWR or PWR conditions.10 SCC of steels containing up to 55 chromium is most frequently observed in caustic and attrate solutions and in media containing hydrogen sul fide.20,21 4.6 Hvdennan Fmheittt nt Hydrogen embrfttlement occurs when a materf al's properties are degraded due to the presence of excessive amounts of hydrogen in a metal's cryst:,i lattice.
This type of damage is a mechanical /envimnmental failure prxess which usually occ urs in combination with residual or applied str esses.
Hydrogen embrittlement is observed most often in plastica 11y deformed metal:
or alloys in high pressure hydrogen environments and is characterized by ductility losses and lowertnc of the fracture strwss.
The only likely i
source of hydrogen in the area of the lower pressurizer head is hydroger.
produced during any cormston reaction.
However, corrosion tests en low alloy steel in deaerated boric acid solution ind hated that the maximun hydrogen present in the steel from corrosion was less than 2 ppn and did no*:
increase with total corrosion on time.17 Hydrogen embrittlement would, t
therefore, not be a probable concern for the ANO-2 pressurtzer.
4.7 3"==a rv of carmaten cana1da rations l
The information presented in the preceding sections indicates that low alloy steel is subject to negitstble general and galvante corrosion in reactor
51-L168554-00 PAGE 11 CF 16 coolant under deaerated conditions.
Under aerated conditicns, corrosion rates up to 9 mpy may be possible; however, the environment in the pressurizer would only be aerated during spectal shutdowns.
The proposed repair results in an extremely ttght crevtce where negl1gf ble ingression of primary water can occur.
The crevice cerrosion rates would be stallar to predicted general corrosion rates if water ingression does occur; however, after some time the cormston products would build up to plug the crevice.
Hydrogen anbrittlement or stress cormston cracking are not a concern in PWR primary coolant.
5.0 *:-- rv mad cane 1u<<ons The proposed interim repair of the ANO-2 pressurtzer leak includes weld repair of cormston damage and insertion of a low alley steel plug to produce an interference fit.
This plug will be welded to the pressurizar shel l 0. 0.
The proposed repair produces an extremely tight crevice preventing the ingression of primary water.
If some water should enter the crevice, corrosion rates should be minimal, and a plugging mechanism should eliminate further water ingression after some time.
No other corroston concerns are evident as demonstrated by both plant and laboratory experiences.
I The' proposed repat r of fers a sat I sf actory interim solutton for the pressurizer leak.
l l
l
.__.. _. - ~ _ -... _ _ _ _
51-1168554-00 PNiE 12 0F 16 Table 1 Typical Water Chemistry Guidelined (Reference 1 and 2)
Shutdown Subcritical/Startup Critical t< EDF1 t< EDF1 L-M C1
< 0.15 ppe
< 0.15 ppe
< 0.15 ppe F1
< 0.15 ppe
< 0.15 ppe
< 0.15 ppm Otssolved 0 aerated 1.0.10 ppm
< 0.10 ppe B(typical) 100-13,000 ppm 100-13,000 ppm 100-13,000 ppa pH 4.6 - 8.5 6.4 - 7.8 H
? 15 ppm 25 - 50 cc/kg Hp
~ < 0.1 ppe at all times.
\\
e
- A.w.
9*
s
51-1168554-00 PAGE 13 0F 16 MnTER G.Q % q
%MT gge c.tw na.,
V SAs35450L(
6*V5 tcweitmAp WELO l
Corer.escu m
1 e
u%_
b swra
51-1168554-00 PAGE 14 0F 16 7.0 Rafa raneam 1.
IE Information Notice No. 86-108:
" Degradation of Reactor Coolant System Pressure Boundary Resulting From Boric Acid Cortzsfon," USNRC, December 29, 1986.
2.
IE Information Notice No.86-108, Supplement 1:
" Reactor Coolant System Pressure Boundary Resulting From Boric Acid Corroston," USNRC, Apr11 20,1987.
3.
"PWR Primar/ Water Chanistry Guideltnes, "EPRI NP-4762-SR, Electric Power Research Institute, Palo Alto, Ca11fornia, September,1986.
4.
J. H. Hicks and E. D. Yochetm, " Water Chemi ts ry Manual for Duke-Type Plants " BAW-1385 Rev.
3, Babcock & Wticox Company, Lynchb urg, Virgint a, Apr11,1980.
S.
G. D. Whitman, G. C. Robinson, J r. and A. W. Savolanen, "A Rev few of Current Practice in Design, An al ys is, Materials, Fabrication, Inspection, and Test," ORNC-NSIC-21, Oak Ridge National Laboratory, Oak l
Ridge, Tennessee, December,1967.
6.
D. C. Vreeland et al.,
" Corrosion of Carton and Low-Alloy Steels in Out-of-Pile Boiling-Wate r-Reactor Envi ronment," Cerrosten, 17(6):
269t-776t (June 1%1).
7.
D. C. Vreeland et al., "Corrusion of Carbon Steel and Other Steels In Simulated Botling Water Cny t s unneen t:
Pnese II,"
car ros t en_,
19(10):
368t-377t (October 1%2).
8.
H. H. Uhlig, corrosion and carrosion Centrol, Wiley, New York,1963.
4
-,-,n, n.-,--.,,___,,-e-
51-1168554-c0 PMiE 15 0F 16 9.
H. R. Copson, " Effects of Velocity on Corruston by Water,"
Ind. rno.
Chaa., 44:1745 (1952).
10.
D.
C.
Vreeland, " Corrosion of Carbon Steel and Low Alloy Steels in Primary Systems of Water-Cooled Nuclear Reactors," paper presented at Netherlands-Norweigen Reactor School, Kjeller, Nonray, August 1963.
11.
W. C. Pearl and G. P. Wo2adlo, "Corriston of Carbon Steel in Staulated Boiltng Water and Superheated Reactor Env f rwnments," Carremf on, 21(8):
260-267 (August 1965).
12.
O. E. Tackett, P. E. Brown, and R. T. Esper, " Review of Carbon Steel Corrosion Data in High Temperature, High Purity Water in Dynamic Systems," USAEC Report WAPD-LSR (C)-134, Bettis Plant, Westinghouse Electric Corporation, Oct. 14, 1955.
13.
D.
J.
DePaul. (Ed.), corrosion and waar Handhook for Water cool a_d_
Rameters, USAEC Report TIC-7006,1957.
14.
W. E. Ruther and R. K. Hart, " Influence of Oxygen on High Temperature Aqueous Corrosion of Iron," correston.19(4):
127t-133t ( April 1963).
15.
E. Howells and L. H. Vaughan, "Corruston of Reactor Materials in Boric Acid Solutions," RCE-1086, Babcock & Wticox Company, Allt ance, Oh to, August, 1960.
4 16.
Yankee Atomic Electric Company to United States Atomic Energy Commission, " Evaluation of Yankee Yessel Cladding Penetrations,"
License No. OFR-3 (Docket No. 50-29), October 15, 1965.
. - -... ~..
51-1168554-00 PKaE 16 0F 16 17.
"Absorptt'on of Corrosion Hydrogen by A3029 Steel at 70F to 500F," WCAP-7099, Westinghouse Electric Corporation, Pittsburgh, Pennsylvania, Decembe r, 1967.
18.
M.
Fe rgu son to L.
E.
Johnson, " Examination of 24-inch Tube Sheet Assembly from the 37-Tube OTSG,"
LR:68:2218-05:1, Babcock & Wilcox, Alltance, Ohio, January 18, 1968.
19.
R. H. Emanuel son, et. al, to R. C. Pittman, "Results of the Operation and Examination of A 19 Tube Model Boiler Damaged to Simulate Crystal River 3-8 Steam Generator," LR:81:5267-05:01, Babcock & Wilcox, Alltance, Oh1o, October 26, 1981.
20.
H. H. Uhlig (ed.), corrosion Hardbook, p.125, Wiley, New York,1948.
21.
G. C. Shvartz and H. M. Kristal, corrosion of chamfem1 Annaratus, pp.
53-70, Chapman and Hall, London,1959.
4 l
.B
- r-
ATTACHMENT C EVALUATION OF LOOSE HEATER PARTS DSG-87-096
}
l
Int:r:ffico C rrarpondanao agggggSTIONfENGINEERING To:
Mr. J. A. Walker Design Services cc:
Mr. J. P. Cook DSG-87-096 Nr. W. J. Heilker
'Mr. F. P. Hill May 3, 1987
Subject:
Evaluation of Loose Heater Parts in ANO-1, Unit 2 Pressurizer
'The pressurizer heater failure at Arkansas I, Unit 2 has resulted in loose parts from the heater within the pressurizer.
The consequences of permitting these loose parts to remain within the pressurizer are analyzed herein. Potential loose parts from failed heaters are:
1.
Inconel 600 end plug - approximate dimensions.625 diameter x 1/2" 2.
Sheath material, Inconel 600 - Sheath is 7/8 00 x.125" wall
('
3.
Magnesium oxide core - approximate dimensions.55" diameter x 55" long (prior to failure) 4.
Magnesium oxide insulation 5.
80-20 Ni-Cr resistance wire 6.
Nickel clad copper conductors These materials are rather inert and will not combine or react with any known material or substance in the reactor coolant system. The magnesium oxide is an insulator and as such an inert material.
It will exist as suspended solids in the reactor coolant but the small quantities will make its presence rather insignificant. Should a small amount adhere either to a steam generator tube ID or the fuel roads, it will not impair the heat transfer characteristics because of the limited quantities of magnesium oxide that is present in the reactor coolant.
The exposed copper from the electrical conductors will slowly (1/2 mil / year) corrode but this will also be insignificant due to the small amount of copper present in the pressurizer.
The flows in the pressurizer through the surge nozzle are small and will not create vibratory motion that could wear ana/or impact upon and damage adjacent heaters.
In all likelihood loose parts will settle in the bottom head of the k
pressurizer and thus be shielded from the flow through the surge screen.
~
E Mr.
J.~ A. Walker DSG-87-096 p'
May 3, 1987 4
i Loose parts could also settle on_ top of support plates. Flows in this region are too small to move parts of any significant size.
The,. calculated fV _ pressuq psi.in the cylindrical portion of the pressurizer is calculated to be 1.8 x 10
' This'.was calculated using the specification surge line flow of 30-lb,/sec.for:
i plant, loading, unloading and operational step changes.
~
[
- The'flowthroughthesurgenozzlemustpgssthroughasurgescreencontaining o
~792 holes at 1/2" diameter each. TheJV pressure created by the flow through these flow holes is 4. psi.
Ple e n te that this flow quickly dissipates.and 2
3 quickly approaches the limiting V value of 1.8 x 10 psi. Furthermore the surge screen discharges fluid i :the horizontal direction in a cylindrical pattern and any pieces near the screen are (due to gravity) beneath the. surge screen.
It is therefore concluded that large pieces will not be moved about in the pressurizer as a result of the surge-flow.
i
' Although unlikely, it is-theoretically possible for very small pieces to exit the pressurizer through the 1/2" diameter holes in the surge screen. These
! (L pieces could'then migrate through the primary piping and into the hot leg side of the_ steam generator. Again it is theoretically possible for the pieces to 4
travel through the heat-transfer tubes and into the cold leg plenum of the steam generator and from there to the reactor vessel. This is a rather torturous path for the loose pieces and in all likelihood the small pieces would be reduced in size from random abrasive action as the parts travel through the primary coolant loop. This abrasive action will eventually obliterate the pieces to harmless particles and thus add to the general' sludge -
?
inventory of the primary coolant loop.
1-A further investigation was made to determine whether pieces lodged against j
the surface of an operating heater could cause hot spots and a resulting heater failure. The loose pieces are of irregular shape originating from a cylindrical rod and it is therefore difficult to postulate anything but line i
contacts between loose pieces and heaters. This will, therefore, not prevent steam bubbles from escaping from the 00 of the operating heater and eliminate the concern for hot spots. This condition is no different than that existing between the heater support plates and the heaters.
5 There are some concerns that loose pieces in the bottom head of the l
pressurizer could impair the operation of the instrument penetrations in the head. This has probably been the case since a modification to these penetrations were required and made at the previous refueling outage. The j
modification allows the penetration to extend 5" above the bottom of the head, thus avoiding debris on the head surface.
Loose pieces could still migrate (fall) into the instrument line and impair their operation. This might not be i {-
a concern if the lines could be "back flushed" should foreign objects enter i
)
e e-r--
- w
,m - w -r y, w,y-
.,,---..,ww-.,,
,,--,,,,-,+-m--a,,~w--,,.--.
---.--m--,-e-,-.--,-..
-,cy,.,--e,,m-,.w,-.u-,--,.--m.
i o
Mr. J. A. Walker OSG-87-096 May 3, 1987 the nozzles. As an alternative to insure'against loose pieces falling into the nozzles it is suggested the present standpipes be replaced with standpipes with closed ends and small holes drilled around the periphery of the standpipes.
p'Tne pressurizer has been operating for some time with debris present and
, 'without any visible or postulated damages to the pressurizer or the reactor coolant system.
It can therefore be concluded that continued operation with debris <in the pressurizer will not result in any harm to the components or the pl. ant.
m J. H. Sodergren JHS/sg
(-
i l
l e
___ _.